`vtight` is not a recognized option for gbsa grid.

[TyBalduf]

Describe the bug
xTB doesn't recognize vtight as an option for the xTB, even though its listed as an option in the documentation. It actually accepts verytight. Some other keywords accept both variations of this string to mean the very tight option, while gbsa only accepts the specific form

To Reproduce
Run xTB with implicit solvent and an input containing

$gbsa
  grid=vtight

it should only use the default 230 points per atom.

Expected behaviour
Runs with tighter Lebedev grid (2030 points per atom).

[chselz]

As this vtight keyword only appears in the man page and not in the readthedocs documentation, I think the most simple fix would be to change the man page, however we could also make it possible that both keywords are recognized. What do you prefer?

[TyBalduf]

I'd be in favor of adding vtight similar to how opt level accepts either.