Just to clarify... segwit would likely result in the tx level equivalent of 1.75-2MB blocks. Only if every transaction was multisig would the 4MB equivalent be reached (translation: it won't).
Precedent that hard forks are "bad and scary" remains firmly in place, to be rolled out again the next time we have this argument.
This is from two weeks ago, so I wouldn't hold your breath waiting for a market reaction.
That pretty much sums the whole discussion in here.
There was very little or no market reaction to the last bitcoin workshop meeting after which that news was released.
I expect more of that discussion to repeat in February, as there is the next big event (in San Francisco) coming:
https://bitcoin.org/en/events