Ethereum has pulled off several hard forks, which they ran within weeks of finalizing the software. There's no reason Bitcoin couldn't do the same, especially for non-contentious forks.
Absolutely, and Bitcoin sort of had too. That knowledge isn't very helpful for managing changes today unfortunately. Today people actually buy stuff for Bitcoin and it's not very helpful to ask all economic activity to cease until a fork has settled down.
None of this is very relevant to the uptake of segwit transactions however, unless as a what-if scenario had segwit been done as a hard fork instead. That was the original intent of the proposal, if I remember correctly, and there was a lot of discussion at the mailing list exactly how this should best be deployed. Looking back at it, it is quite clear it could have been deployed quicker had another mechanism been chosen.
None of this is very relevant to the uptake of segwit transactions however, unless as a what-if scenario had segwit been done as a hard fork instead. That was the original intent of the proposal, if I remember correctly, and there was a lot of discussion at the mailing list exactly how this should best be deployed. Looking back at it, it is quite clear it could have been deployed quicker had another mechanism been chosen.