Does quoting # work when not in extended mode?

It behaves the same in extended and normal mode: https://3v4l.org/f1eg0

Quoting documentation:

For example, if you want to match a "*" character, you write "\*" in the pattern. This applies whether or not the following character would otherwise be interpreted as a meta-character, so it is always safe to precede a non-alphanumeric with "\" to specify that it stands for itself.

I'm somewhat concerned regarding the potential BC break. What if existing code is already properly escaping the #? E.g. preg_match('~^(' . preg_quote('hello\#world', '~') . ')\z~x', 'hello#world') would return 0 with this patch applied.

Your code is not working now as well because \ gets escaped anyway: https://3v4l.org/O8B5o (without x mode to avoid broken regex)
You can't properly escape # before preg_quote(), only afterwards by e.g. addcslashes(preg_quote(...), '#').
Such code would be broken by this PR I suppose. Yet this is still a bug that should be fixed at some point.
BC concern is legit, although it's a rare use / edge case.
Absence of escaping # could even be considered a security risk in some scenarios, because you can break the the pattern this way and entirely bypass the validation (i.e. bypass \z).

You can't properly escape # before preg_quote(), […]

Of course, you're right.

Anyhow, it might be worthwhile to discuss this issue on internals to reach a larger audience.

@Majkl578 please link back to internals discussion (via externals preferably) for future reference ...

Merged as 8423534 into master. Given that we have already added characters to preg_quote in the past and given the potential security vector, I think it's reasonable to introduce this in the next minor. If this breaks someones code, they have the next year to report it...