This may seem nit-picky, but I think there is benefit in being precise wherever possible. For example, if when filling out the questionnaire, one checks the "Momentum" box, there ought to be a way to measure the momentum-ness, or the word "Momentum" could just as well be replaced with "Pixie dust."
The Q risk model is described here:
Therein, we have definitions for:
Presumably these are the definitions used in the smart beta questionnaire.
We still need definitions for:
Mean-reversion (same as short-term reversal in the risk model?)