The reason is that I don't find your answers convincing, nor rigorously analyzed or presented, for the most part. That even applies to Peter R, in terms of many of his answers on this. His paper was good but it only addressed a small part of the larger set of questions.
then i'm sure you'll accept the fact that i find your fears even less rigorously presented.
at least when i throw up all those graphs from Tradeblock, Chain.io, statoshi.info, i analyze and attempt to explain what i see in the data. and what i see during stress tests, is a network that is being artificially constrained in that there are no massive full node failures as a result of their unfortunate wasted needs to validate inflated mempools, i see no massive orphaning (that's b/c all pools are processing 1MB blocks together), i see no delays in the avg 10 min block processing, i see no problems in the storage capacities of any of my full nodes, and i hear no complaints from miners.
all i hear is complaints from users who get unconf stuck tx's and high fees and wonder what the hell is going on? sure, there's a small subset of us that know how to get thru but that is a small minority and totally ignores all those newbies who could be coming onboard to Bitcoin if not for the problems the congestion caused. what a wasted opportunity we had a month ago during the Greek peak crisis.
and here's the thing. i think all of us have access to the same data all the core devs have access to for the most part. even the technical simulations they rarely do. so really it boils down to interpretation of what we see going on and i see artificial constraint.