I reviewed the patch and the speed test attached. 
Well, yes, the patch does achieve a certain level of speed improvement but I also saw that in cases when the quoting is really required (special characters, the current stdlib is faster). 

The speed improvement proposed is really not a convincing enough reason to replace the code of the current quote. rhettinger's comment in this issue might taken as reference to work out if there are any statistical evidence (:)?) that if the frameworks use it, would they benefit.

Few more points.

- quote is a generic function, checking for empty string, returning it  and marking it speed improvement, really is not a improvement (practically the person using the quote can avoid it too and thereby increase the speed).

- Some framework might try this , like django or zope for e.g and see if its really a worthy case in using re to check for chars to be quoted and quote only those (and see if there is any improvement).

- Might provide this as an activestate recipe and see if there any takers (again for reasons of use-case)

It is correctly a low priority one, or it can just be rejected too.

BTW, there is a feature request on quote support unicode char with patch, when that is in place, I think this request may suffer further.