I think that the simplest way to understand DMS.MINING is that the correct price is what the market thinks 1 mh can earn in about a year (plus a safety margin). That time frame is very important because we can say pretty reliably that over a substantial portion of a year difficulty will probably not trend downwards or stay the same. If we extend the timeline further, say to two years, that would be even "safer". If we reduce the calculations and use 6 months, or 3 months, then we run the risk of collapse because DMS would be entitled to ever-increasing dividends out of a fixed pot. Unless new buyers of DMS.PURCHASE come in to provide more capital, I don't see how the fund could avoid catastrophic failure.

So unless I misunderstood, the single most important feature defining the correct value of DMS.MINING is the period of time specified in the contract. If it were calculated based on a 6 month return, we'd have much higher yield per mh, but higher risk of catastrophic collapse. If it were calculated based on a 2 year term, we'd be much safer, but would expect much lower yields per mh. This, I think is the key distinction between DMS.MINING and TAT.VM.  If difficulty levels off in the future, TAT.VM has no problem with it, but DMS.MINING could collapse.