The Flexcap idea doesn't make sense to me.  It seems Adam is trying to artificially make it more expensive for miners to publish large blocks.  The reason this doesn't make sense to me is that even without a block size limit it would be more expensive for miners to publish large blocks due to the orphan cost.  Why use "Flexcap" to simulate the natural supply and demand dynamics that already exist?

The fact that larger blocks are more costly to produce is illustrated in Fig. 8 of this paper:



As an example, in the absence of a block size limit, it would cost a miner approximately 100 BTC on average to publish a 128 MB block, assuming a propagation impedance of 7.5 sec / MB.

In a discussion with a 0.5% miner, Adam recently said:

   "Now if there is an excess of supply, price falls, ergo fees will drop to zero basically."

which is not true and demonstrates Adam's lack of understanding of the transaction fee market.  If the block size is not constrained by the protocol, then the fee per kilobyte is governed largely by the orphan cost. The orphan cost is a function of the propagation impedance for block solutions. Raising the block size limit does not affect the propagation impedance. Fees per kilobyte would be largely unchanged; however, since more kilobytes of transactions could be included in a block, the total fees per block could grow higher.

TL/DR: Flexcap reinvents the natural fee market.