But, as it stands now, even 8MB blocks might be risky at the moment
Just noticed the disconnect between your message and your username.
I appreciate your attention to details for I value that just as well.
My username suggests the next bandwidth target for the idea of a static limit on block size, but doesn't hint at the time when the adjustment needs to be made. I argued against many other more sophisticated scalability proposals with mathematical rigor, but it was when I found myself defending the future limit against the current one, that I realized the obvious contradiction in my line of thinking.
Finding the appropriate moment in time for the transition would become the most challenging and interesting part of Bitcoin's evolutionary path.
I'm probably one of the most rabid '1MBers' around and have been for years. Ironically, I have no real problem with going to 8MB. I only want to make sure that it happens AFTER a workable scaling mechanism to support 'every frappuccino purchase worldwide' has been fielded and proven. Since way back in 2011, 'subordinate chains' has been what I see as the best solution in terms of achievability.
It became clear to me many years ago just on simple math that even at 1MB, jump-starting a full verifying node would be most practical by transferring physical media. As long as the block size remains such that the blockchain can be passed along on a single and affordable consumer-grade storage device, I'm mostly happy.
If certain critical functions of 'native Bitcoin' can be effectively 'carried along' by those operating a purpose-oriented sidechain (for instance) then the pools of potential infrastructure support for native Bitcoin could be greatly expanded. That is another metric I use in deciding what scale is acceptable and what scale is to onerous.