Where I struggle is at the beginning, point 1. that looks at the cost of producing the block. I know that it costs to miners to find blocks, but the link of this cost with the block reward is what bothers me. The block rewards are created by software independently of the mining cost.
Sure, you can look at it that way. i.e. The block has value..."because"...rather than the block has value "because the prevailing level of competition to mine it represents the starting value for the block".
However, taking that approach has consequences - namely that half the supply ends up being held at a zero cost base and free markets do like to massacre high margins where they have the option. So you end up with chronic profit-takes from masternode rewards competing with miners for limited fiat liquidity and undermining them (because they can afford to right down to a price of zero).
It doesn't matter which way you model it, the long-run behaviour (in the absence of massive added service value to justify the MN rewards) tends towards chronic loss of marketcap share.