This is, however, a crucial problem for Segwit2x: An user would have to install the 2x client if she wants to receive money from the 2x chain and use (send) it. So after some time - some days, some weeks at most if she doesn't use BTC too much - she must install the 2x/Btc1/Electrum2x/whatever2x client if she wants to use that chain.
The question is now how many users will do this. Exchanges and other companies that support Segwit2x will surely try to promote a Segwit2x client, and they could have some success because from what I perceive, most non-technical users are "apolitical" in the block size debate and would use the client that gives them personally the greatest advantages.
This is the age-old conundrum of hard forks and why many believe they will inevitably lead to multiple blockchains in a mature currency like Bitcoin. I wouldn't be surprised if most Bitcoin users don't even know/care about the block size debate and various forks. They've just been running Core or Electrum and will continue to do so. Bitcoin is a real, live economy. Businesses run like clockwork on it with scripts and APIs. Every user manually, simultanously installing a new client? In Bitcoin? Give me a break -- impossible. On smaller and/or more centralized networks, this isn't necessarily true, as we've seen with Ethereum and Monero. But clearly hard forks lead to chain splits once stakeholders really dig their heels in.
The most likely outcome is, in my opinion, that the Core chain will "win" the battle, Segwit2x will be marginalized, and big blockers will switch to Bitcoin Cash definitively. That could have as a consequence that Bcash will be stronger than now, although I don't think it can ever be the leading coin.
I think that's what Bitcoin Cash supporters are hoping for, and why they seem generally supportive of 2X. I suspect you're right. I'm not sure which altcoin chain will win out between 2X and BCH, but I think probably BCH. It's the big blocker's bloatchain of choice.
~
I don't think it's possible to use nLockTime in Electrum. Not in the standard user interface, anyway. I've browsed through Electrum's command line documentation and I don't see any reference to nLockTime, either, so I'm at a loss for how to make it work.
i couldn't find it either. but i remember something about this!
you can always make the transaction and before signing it, manually change the last 4 bytes of the raw transaction (8 hex characters) to your locktime
a little advanced but not hard.
for example the transaction ends with: "....88ac00000000" for locktime of 0. and last block height now is 488503. you set it to 488504 which is 38740700 in little-endian "....88ac38740700"
Wow, you learn something new everyday. I'll experiment with some very small transactions.