I totally agree block size will need to raise through time. It's just wrong to think of that as a scaling solution.
The protocol has this nice automatic difficulty adjustment that takes care of increasing compute performance, a shame there is no automatism to detect how big a lower-end hard drive is currently and adjust block sizes accordingly.
The problem with automatic adjustment of the block size is that we have no way of really knowing what the optimal block size is based on onchain data alone or by any metric. What the block size should be is really down to opinion the way I see it.
I think that targeting a block size in the future that keeps a regular onchain transaction cost at around 250 USD in today's value about right.
Onchain transactions should be expensive, just not too expensive.
reply