fix: correct the signature of .events()

[PietroPasotti]

Fixes #1341

[tonyandrewmeyer]

Thanks! Curious that pyright doesn't complain about this.

[PietroPasotti]

I think it didn't because all attributes that were accessed were coincidentally common between the two types :)
they both talked like ducks

[tonyandrewmeyer]

LGTM, thanks!

The main case seems very straightforward - the code is clearly wanting the bound event - it's even asserting on it. The error message probably works either way, but I'm +0 on this adjustment.

The framework one should indeed be working with bound events - the class has the source, the instance has the bound event.

[benhoyt]

Thanks!