Determine ideal ordering of parsers in `recursiveQ`

[valencik]

Good point!
I tested with termAndWildCardQ last and all tests still passed.
However, these cases are checked in order, so moving it around could have some performance implications.
In the sense that it could be faster or slower, I don't know.
I'm going to open an issue to look into this, and perhaps use our benchmarks to help come up with a principled approach to deciding on this order.

Originally posted by @valencik in #154 (comment)