Compare performance hit of functions called from native Rust vs Torchbear

#217 was very informative

I think something like seeing how many times `checksumdir()` can be called native vs scripted would also be informative

perhaps also `regex`

I want to see how much sense it makes in mid term to keep logic in Lua through Torchbear's bindings. as far as things are going now, I see it this way:

> torchbear balances speed, efficiency, and power. it's the easiest way to quickly build something powerful that works well on affordable hardware.

mid/long-term, I think the libraries will still be there and get better from torchbear making them easy to use. and there will be ongoing rewrites of whatever logic we have in Lua libraries up to native code, like what is partly going on with ContentDB

long-term, I intend to transition away from PUC-Lua but with a very similar syntax and use a more effective runtime in Rust to provide an optimal balance between high and low level programming