I've highlighted the constraints of the network with a larger block size and why a block size that is excessively large isn't favorable for the network. I'm not concerned about your algorithm, I'm concerned about whether you think it is okay for the so called self-regulated system to regulate the block size such that it can potentially take up to minutes to propagate across the network.
I understand your point but if the limit is not high enough then many people won't even be able to spend their coins which is another problem.
Do you think the point of the block size we have today is to introduce scarcity into the block or is it ensure that the blocks are appropriately sized and thereby preventing any issues, pertaining to the security of the network or it's resources? Your self-regulated system is not going to work if it can have the potential to make the network excessively centralized or insecure. There is a very good reason why most of Bitcoin's derivatives don't have a dynamic or an unlimited block size (for which a dynamic block size without any hard cap can also be considered as an unlimited block size).
It is to protect the network and I think we must have a limit. I just think the free market should decide what that limit should be and the miners should adapt accordingly.
Edit: If you think that it is fine for a dynamic block size without any upperlimits, then I rest my case. I don't have anything to add on ontop of what I've mentioned and I'm not a huge proponent of increasing it to meet a level that could be compared to the TPS of a mass adoption.
I believe you prefer a block size increase based on technological progress and I favor a block size increase based on demand. I proposed a way to calculate the block size limit, so far you haven't. I'll briefly explain how my equation could balance supply and demand..
new limit = (current total fee - previous total fee) / current avg + previous size limit
We calulate the excess demand or excess supply relative to the previous period then we calculate the new block size limit to meet a new equilibrium (assuming that the demand is constant).
Could you tell us how would you determine what the block size limit should be in the future?