There's evidences that salvinorin A is byosynthesized from salvinorin B by an acetylation enzyme.
Some species have high level of salvinorin A and low level of salvinorin B. Others the opposite.
This suggests that in the first group the responsible gene is very active and that in the second one it is not.
Salvia recognita is in the first. Salvia potentillifolia and cryptantha are in the second. All this three species have 8 pairs of chromosomes, and are phylogenetically close each other.
I think is very probable that they are cross-fertile, and that their offspring will be fertile too.
I hope that the active acetylation trait of recognita and the high production of salvinorin B of potentillifolia or cryptantha can be inherited by some of their descendant.
Salvia divinorum, sclarea and coccinea have 11 pairs of chromosomes.
Salvia coccinea is not very vigorous, can stand light frost and is phylogenetically close to divinorum. Salvia sclarea is very vigorous, can stand deep frost but is not phylogenetically close to divinorum.
Here also I hope that some offspring can be produced and that will inherit convenient traits.
I'm searching for potentillifolia and cryptantha seeds sources. They hard to find since they are not commercially interesting. I only found one seller at the moment, but the seeds I bought didn't germinate.