Exactly. The grace period should be 6 to 12 months IMO. What Gavin suggested with Classic is a joke (28 days). Businesses can't successfully develop, test and deploy in that time-frame. Even if they could, it would likely end up problematic due to it being rushed. The only way that it *may* work for e.g. <6 months if everyone was running non-custom versions of the software, which we know is not the case.
You know, what I find very unfortunate and saddening is the toxicity towards Segwit. I mean, it's a *wonderful* development that will enable Bitcoin to scale further and better. The usual fork supporters are very negative towards it, while they should be appreciative in the lines of (random generic statement) "Great development! Now we can and should plan a hard fork.". I don't participate in those subreddits, but I do read them for behavioral analysis.
Ethereum hardforked in 1 week. That is the real joke, and nobody seemed to care about that, hence ETC was born as a thorn in their asses for this negligence.
I've seen argumentation using the fact that the Bitcoin 'dominance' percentage is at ~80%.
It's 74.8% now but yes, BTC is worth more than all other coins combined. So many people could try to fork it, and they did (altcoins are forks too essentually) but none of them beat BTC so far.
I think the dominance will shrink later, since new innovating coins could come into existence, but I dont believe on a 1-1 basis any coin could even come close to beating BTC.
So later the crypto market cap could be say 200 billion $, of which 100 billion could be BTC and 1 billion for each succesful altcoin, but hardly any more.