I've gotta say, I'm slightly surprised (in a good way). I mean, I figured people would just point out flaws and "boo" this one into oblivion (and while that's mostly what is happening, it's not to the extent that I imagined it would be).
Here are a few more thoughts (after digesting the responses and doing some thinking of my own):
(*) I think 0.1 is somewhere in the ballpark of a merit-to-post ratio that should be exceeded if someone's trying to grow their rank. I mean, I realize that not everyone can contribute at the same level, and so, I'm trying to find a ratio that I think most every person can hit without giving themselves a hernia (so to speak). If all I wanted was to drive up the forum's signal-to-noise ratio while not really considering where this might leave people who depend on the forum for some amount of their income, then I'd be tempted to set it much higher (like 1.5, which while 15 times greater than 0.1 and completely unrealistic-seeming to most, is still ~6 times lower than my own merit-to-post ratio). The problem with setting it too low (like 0.01, or something) is that we would then have complexified things in exchange for no real improvement to anything. If 1000 posts can be produced while only risking a backslide of 10 merits, then I predict very little will change by implementing this "carry" thing. But if writing 1000 unappreciated posts can end up costing you 100 merits, then that's genuinely behavior-affecting, IMO. As with everything Bitcoin(talk), I'm pro-freedom. As in, if you want to make 1000 posts that the community is effectively "telling you" (by not leaving you any merit) that they don't appreciate, then that's fine, and you should feel free to keep doing that. But, if for some reason that's the posting pattern that you've fallen into and you're still very concerned with maintaining or increasing your "rank", then something like what I'm proposing will encourage you to find a more community-appreciated way to spend your time on the forum. (And when you think about it, what kind of BS user is that, anyway? Someone who's at once saying "I have little to offer" and "I wish to be highly regarded"?)
(*) There would need to be some exemptions, I think. ChartBuddy should be able to post without building up "carry" (though, like LV said, probably the WO crowd would have been happy to keep CB fed with merits). Also, people should be able to post redirection topics without worrying about "carry". Maybe the "Serious discussion" and "Ivory Tower" boards should be exempted. That kind of thing. Because the "carry" database table would be event-style in the same sense that the "merit" database table is (as in, balances in either case are derived by summing over an adjustable record of events, rather than just being stored opaquely), I don't think it's very important to ahead-of-time think of every piece of special casing. Over time, cases can be made for specific things to be exempted (like maybe some of LV's data-dumping threads, for example), and even if those cases are only discovered after they've already caused unwarranted accumulation of "carry", it would be very easy to after-the-fact fix that with retroactive exemptions. (In general, though, if I were making the decisions, which I won't be, I'd lean toward "no" for most exemption requests. There's something very natural to me about letting merit decide whether or not the community appreciates something, and I struggle to think of clear examples of things that definitely should keep happening despite the fact that they don't/can't carry their own weight in terms of receiving merit.)
(*) It's natural for me to consider things in very abstract ways and to keep my mental "wavefunction" from collapsing around concrete details that aren't central to the "shape" of whatever it is that I'm thinking about. But, I know that that's not how everyone thinks, and some people prefer a style of thinking where concrete details dominate their analysis of something. In my experience, it's very frustrating trying to transmit ideas across an abstract-thinking vs. concrete-thinking "impedance mismatch". With the particular proposal in the OP, I've tried to "project" an abstract idea down onto the "concrete plane" so that it can be easily discussed. But, it's important to remember that many such "projections" exist, and there's no real value in picking apart just one of them (it's only really valuable when you've found a problem that would exist in every concrete projection of an abstract idea). Basically, what I'm saying is, I'm not very attached to any given detail in the OP, and I only really have a sense that something like what I've suggested could be made to work. For example, maybe it should instead work like: If you've received any amount of merit during the previous 14-day "activity" window, then all the posts you make in the current 14-day window won't attract any "carry" (all that that changes is the specific logic driving when the system does and doesn't generate a "carry transaction", but the scaffolding remains: a subtractive term in the rank calculation that's kept track of with an event-style database table).
(*) Back when rank used to be post-count-only, and things were adjusted to be based on "activity" (in June of 2013), many posters were all like "Why change this? FFS, I liked how it was before!". Then, when an additional term was introduced (in January of 2018), again, many posters were all like "Why change this? FFS, I liked how it was before!". So, it's inevitable, IMO, that any rank-hardening adjustment made at this point, even one that manages to "thread the needle" just so and long-term improve the forum's signal-to-noise ratio without causing too many casualties in that pursuit, will, at the time of its proposal or implementation, be met with some amount of, wait for it... "Why change this? FFS, I liked how it was before!"
Here are a few more thoughts (after digesting the responses and doing some thinking of my own):
(*) I think 0.1 is somewhere in the ballpark of a merit-to-post ratio that should be exceeded if someone's trying to grow their rank. I mean, I realize that not everyone can contribute at the same level, and so, I'm trying to find a ratio that I think most every person can hit without giving themselves a hernia (so to speak). If all I wanted was to drive up the forum's signal-to-noise ratio while not really considering where this might leave people who depend on the forum for some amount of their income, then I'd be tempted to set it much higher (like 1.5, which while 15 times greater than 0.1 and completely unrealistic-seeming to most, is still ~6 times lower than my own merit-to-post ratio). The problem with setting it too low (like 0.01, or something) is that we would then have complexified things in exchange for no real improvement to anything. If 1000 posts can be produced while only risking a backslide of 10 merits, then I predict very little will change by implementing this "carry" thing. But if writing 1000 unappreciated posts can end up costing you 100 merits, then that's genuinely behavior-affecting, IMO. As with everything Bitcoin(talk), I'm pro-freedom. As in, if you want to make 1000 posts that the community is effectively "telling you" (by not leaving you any merit) that they don't appreciate, then that's fine, and you should feel free to keep doing that. But, if for some reason that's the posting pattern that you've fallen into and you're still very concerned with maintaining or increasing your "rank", then something like what I'm proposing will encourage you to find a more community-appreciated way to spend your time on the forum. (And when you think about it, what kind of BS user is that, anyway? Someone who's at once saying "I have little to offer" and "I wish to be highly regarded"?)
(*) There would need to be some exemptions, I think. ChartBuddy should be able to post without building up "carry" (though, like LV said, probably the WO crowd would have been happy to keep CB fed with merits). Also, people should be able to post redirection topics without worrying about "carry". Maybe the "Serious discussion" and "Ivory Tower" boards should be exempted. That kind of thing. Because the "carry" database table would be event-style in the same sense that the "merit" database table is (as in, balances in either case are derived by summing over an adjustable record of events, rather than just being stored opaquely), I don't think it's very important to ahead-of-time think of every piece of special casing. Over time, cases can be made for specific things to be exempted (like maybe some of LV's data-dumping threads, for example), and even if those cases are only discovered after they've already caused unwarranted accumulation of "carry", it would be very easy to after-the-fact fix that with retroactive exemptions. (In general, though, if I were making the decisions, which I won't be, I'd lean toward "no" for most exemption requests. There's something very natural to me about letting merit decide whether or not the community appreciates something, and I struggle to think of clear examples of things that definitely should keep happening despite the fact that they don't/can't carry their own weight in terms of receiving merit.)
(*) It's natural for me to consider things in very abstract ways and to keep my mental "wavefunction" from collapsing around concrete details that aren't central to the "shape" of whatever it is that I'm thinking about. But, I know that that's not how everyone thinks, and some people prefer a style of thinking where concrete details dominate their analysis of something. In my experience, it's very frustrating trying to transmit ideas across an abstract-thinking vs. concrete-thinking "impedance mismatch". With the particular proposal in the OP, I've tried to "project" an abstract idea down onto the "concrete plane" so that it can be easily discussed. But, it's important to remember that many such "projections" exist, and there's no real value in picking apart just one of them (it's only really valuable when you've found a problem that would exist in every concrete projection of an abstract idea). Basically, what I'm saying is, I'm not very attached to any given detail in the OP, and I only really have a sense that something like what I've suggested could be made to work. For example, maybe it should instead work like: If you've received any amount of merit during the previous 14-day "activity" window, then all the posts you make in the current 14-day window won't attract any "carry" (all that that changes is the specific logic driving when the system does and doesn't generate a "carry transaction", but the scaffolding remains: a subtractive term in the rank calculation that's kept track of with an event-style database table).
(*) Back when rank used to be post-count-only, and things were adjusted to be based on "activity" (in June of 2013), many posters were all like "Why change this? FFS, I liked how it was before!". Then, when an additional term was introduced (in January of 2018), again, many posters were all like "Why change this? FFS, I liked how it was before!". So, it's inevitable, IMO, that any rank-hardening adjustment made at this point, even one that manages to "thread the needle" just so and long-term improve the forum's signal-to-noise ratio without causing too many casualties in that pursuit, will, at the time of its proposal or implementation, be met with some amount of, wait for it... "Why change this? FFS, I liked how it was before!"