I'd be even more curious about the computation used to determine if a given set of sides and angles tessellates, and how long that computation takes. Approaches to enumerate possible expressions don't seem that far-fetched, if the resulting possibilities can be evaluated quickly.