long-term large change: make >1 of a thing faster

[sckott]

Throughout the package we generally have each function that generate some fake thing make 1 of them. And then if a user wants more than 1 then they run it many times via for, replicate, apply, etc. This is probably not the fastest approach.

Maybe we make all methods accept a n parameter, with default n=1. Internally within each method we would use the fastest approach for whatever is being generated. This would add more complexity of course.