Use xtb for the WBO

[jthorton]

At the free energy workshop, it was mentioned that Cresset is now using xtb to calculate the WBO for fragmentation as they do not have access to openeye. This seems to be more reliable than the use of ambertools SQM and even fixed some fragmentation issues they saw with large ligands such as those in the ptp1b (I think this was the set). This would probably need to be a toolkit wrapper to work but I am not sure if we would want to support that fully although the functionality should be quite minimal similar to the ambertools wrapper.

[mattwthompson]

I'm pretty sure ToolkitWrappers should work even if they're written externally. I've never actually used XTB but if there's a small (1 or 2?) number of calls that need to be wrapped, it seems doable.

[mattwthompson]

This would be fun to implement, if it'd still be useful. This might be where to start?