@Roger_Murdock
It's whether a "small-network, big-block Bitcoin" would be worth more than a "large-network, currently small-block (but with latent capacity for future block size increase) Bitcoin."
Well that's why I said "even when blocks were full and fees were soaring." I was trying to paint a scenario where for some odd reason the miners seem to be nearly committing suicide, which of course we would never expect them to do, at least in theory* (inasmuch as they are a proxy for the economic majority; however, we also expected them to up the blocksize a while ago). In that case the network effect could be outweighed by the sheer brokenness of the small-block network.
I also consider the
argument about future valuation (including future network effect) being baked into the current price.
To better contextualize my comments, I think the idea of changing the PoW algo makes a bit more sense for Core than for Classic, because in the case of Core they have the settlement-coin vision, which basically takes on the investment proposition of an altcoin that might be called SwiftCoin or something, and like you say they can always increase the blocksize later if needed (this has always been Core's strongest argument, IMO; Greg and Luke even think a blocksize increase could be rolled out in 2 weeks if necessary).
@Mengerian
But "we" would also have to buy the new equipment to get started, so why wouldn't that also be a waste of money?
Because we wouldn't get 51% attacked for the reasons I mentioned, and (we believe) our coin would eventually be very valuable even if the market hasn't realized that yet. Just like Bitcoin, it's only profitable to buy the equipment to mine it if you believe in it, not to attack it.
[doublepost=1453308155,1453307234][/doublepost]Or look at it this way: it's at least conceivable that Core is correct that there are trade-offs between blocksize and security, so that it may make sense - at some blocksize at least - to split Bitcoin into SettlementCoin and PaymentCoin. I happen to think that blocksize would be very large, if there even is any such blocksize, while Core seems to think it's right around 1MB.
In that case, the market would support a fork where the minority and majority sides both survive, because it would be rational. Likely their market caps would differ quite a bit, and they'd want different PoW algos to avoid conflict.
I think some are taking my comments as, "Classic
should change the PoW algo if it falls into the minority." I would recommend against that at this point in time, because it's not nearly urgent enough (and likely never will be since Core will capitulate or miners will change their tune, but who knows). In that case, Classic should be delayed until it is more urgently needed or some more miner-favored client is released, and in fact that is what the 75% cutoff ensures. I'm a lot closer to saying Core should do that *if* they want to take a super-hardline stance, which of course I will say is a bad idea but it is their choice.
TL;DR: Changing the PoW algo is the logical thing to do if you really want to take the ultra-hardline stance, even against the mining majority. The question is whether it is ever useful to take such a hardline stance, and the answer is quite likely no. (However, this is still a relevant point since there
are people in Core at least
saying they want to take such a hardline stance. I feel obligated to ensure they are welcomed to do so; it benefits us if they do anyway.)
*
If we really take this seriously, why are we even bothering to post and convince anyone? 
