IMO all will agree to this IF
1. It is the ONLY quick straight fwd solution
2. there is enough proof due to testing incl all preripheric appl
3. nobody is left out / no ugly side affects come up with
As far as a 2 MB block size proposal is concerned:
1) It isn't a quick solution as HF's were never meant to be deployed as proposed recently by Gavin (grace period & consensus threshold are too low).
2) Adequate testing in some areas (maybe somebody has a source)
3) They are left out if they don't upgrade in time.
Its not clear to me that raising the blocksize wouldn't cause massive amounts of node centralization. I think it's going to be really hard to run a node. We need data before doing this, and the research done so far points at centralization of nodes.
Exactly. We need much more data in this case. A lot of people tend to say that it won't happen because storage is cheap and internet is fast already, but I tend to disagree because there aren't even that many people who run nodes these days.