The only way to make software secure, reliable, and fast is to make it small. Fight Features. - Andy Tanenbaum 2004 Oh yeah? So adding complicated stuff like sidechains and lightning network is making bitcoin code smaller?
To the extent those build on top of Bitcoin they aren't adding features to it or making the the core code bigger. If they require changes to the core code or adding new features to it, that's a different matter that needs to be considered carefully.
But if the claim is that sidechains are the 'solution' to the 'problem'. Then you are saying they are part of bitcoin [the ecosystem] whether they are part of core or not.
Increasing block size does not add features *and* it 'solves' the 'problem'. Without introducing any other layers of complexity, or attempting to artificially manipulate the fee economy that is growing organically just as it was always intended.
What frightens me is that the whole thing seems to have turned into a pissing contest.
The fee market isn't a problem right now, the block size is. So why are people trying to pre-emptively fix the fee market in an ass backwards way to address the block size problem.
Nobody is claiming to have predetermined that sidechains (or LN) are necessarily the ENTIRETY of the solution to the scaling problem, only that they are potentially part of it. Bitcoin is an ongoing bleeding-edge experiment, and we are working with intuition, educated guesses, hypotheses, and prototypes, not off-the-shelf kit in neat little boxes.
Increasing block size does add features (and/or bugs), in the form of higher tps and whatever concomitant other new auto/adaptive regulation mechanisms come along with the eventual solution. 100k max tx size is probably only the first such
required adjustment.
And thus, the 100k max tx adjustment neatly destroys your claim that larger blocks do not introduce additional complexity. More is different; the dose makes the poison...
The Red Queen interpretation, whereby we must change Bitcoin ASAP for the sake of keeping it the same, is absurd. Moving the tx supply curve with the goal of controlling the range where it intersects that of demand is prima facie centralized market manipulation. As long as tx fees are absurdly underpriced, in terms of their cost and what users are willing to pay, the fee market is completely broken. Bitoin's 'free sample/loss leader' viral marketing campaign phase ended with the emergence of omnipresent 'cosmic background spam.'
The "pissing contest" which "frightens" your delicate sensibilities is exactly the "fight" to which Tannenbaum exhorts us, because the "adversarial process is valuable in assuring [features] do not compromise security or reliability."