One methodology is to apply Google search statistics to their terminology:
Alas, most of their terms typically don't have enough volume to allow the use of Google Insights for Search. However I can do this:
In other cases, I'll have to use static figures.
|"memetic drift"||9,230||"cultural drift"||53,500|
Of course, science is more than just a populatity contest - and there are about a million other ways in which this methodology is limited.