optimize vec.extend(slice.to_vec()), take 2

[the8472]

Redoing #130998
It was reverted in #151150 due to flakiness. I have traced this due to layout randomization perturbing the test (the failure reproduces locally with layout randomization), which is now excluded.

[rustbot]

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[tbu-]

Does it make sense to add an optimization that breaks with layout randomization?

[the8472]

Layout randomization is primarily for debugging (it pessimizes struct layouts after all), and the optimization is nice-to-have so I think it's ok.
At worst it indicates that the optimization is brittle, but even with randomization it still worked for the Vec->Vec case, just the String->String case broke for some reason, not sure why.