Jameson, thanks again for sharing the talk — I appreciate how deeply you’ve modeled this idea and that you’re grounding it in existing altcoin experiments.

That said, I think the analogy breaks when we scale to Bitcoin’s adversarial environment. Large industrial miners on Bitcoin are closer to oil and gas giants or big construction firms: they have the capital to exploit cyclical dynamics, flooding the system when it suits them and constricting it when it maximizes their margins.

In a variable block size system, even with upper and lower bounds, you introduce a flex point that large miners can game — expanding block size to squeeze out small players, then contracting it to capture profits when the field is cleared. This kind of cyclical exploitation simply doesn’t arise at the same scale on smaller chains, so modeling from them might understate the centralization risk.

And you yourself acknowledge the political and technical difficulty of this kind of transition: not only is it hard to know in advance what the “right” parameters are, but the chances of getting it right the first time are extremely low. So the risk of getting it wrong — and then needing to renegotiate consensus to fix the consequences — is enormous. Once you’ve introduced a dynamic system, the very actors benefiting from the new arrangement will have every incentive to resist future corrective changes. You’d be opening a long-term governance battle that could lock in new power asymmetries.

IMO keeping block size fixed forces miners to compete within a predictable, constrained resource, avoiding the oscillating game-theory cycles that favor deep-pocketed actors, and reducing the need for repeated political fights over core consensus parameters.

And for what it’s worth… the more I explore the differences between the potential of OP_RETURN vs. how witness data is currently used, the more I see the logic behind the shift and its potential to at least temporarily solve the demand problem. This includes practical benefits for node runners: OP_RETURN driving out witness data use is 4x more efficient… so smaller blocks on net is a plausible vector because OP RETURN is more useful without third party services.

I still think starting with lower limits and titrating up as innovations prove their utility is the better systems design approach… but that ship sailed with the original segwit upgrade. I also maintain it would give Core more credibility if they tried a lower limit on both before going balls to the wall max witness/OP_RETURN but I’m Johnny-come-lately to this game so who the fuck am I to say? This whole situation is showing “grass roots” users the true nature of Bitcoin- that it listens to the ECONOMIC majority of nodes. This truth is hard to swallow.

It’s sad we ended up in this place because Bitcoin loses a lot of its ideals for a social good, but maybe that’s just reality and human nature… something we may grapple with for eternity.

Appreciate you brother and the educational path you’ve led me down here.