Hi B,
Being that they have different second(species) names I think crossing is probably low. I'll try to explain.
If you are growing E. purpurea(regular purple coneflower) and E. purpurea 'White Swan', they will cross being they are both E. purpurea varieties. Most of the popular cultivars are E. purpurea. A baby I saved of a E. purpurea and E 'White Swan' cross came out paler pink than the regular.
But if you are growing E. atrorubens and Echinacea pupurea 'White Swan' and E. paradoxa, you should see no crossing of seeds.
You do have to be careful though. Some species of echinacea are not thought to be true species and can cross. An example of this is Echinacea tennesseensis. It is thought to be a sub species of E. purpurea, and if so, cross breeding would readily happen between E. tennesseensis and E. purpurea.
Did all of that make sense? lol. I hope so.
As far as keeping seeds pure if you are growing two crossable cultivars, you need to bag blossoms. They are probably outbreeding flowers meaning you need insect pollination to get viable seed. In that case, you need to bag a few blossoms, uncover them when full of pollen and pollinate them yourself(a small brush works if you can't rub the blooms together) then recover them immediately to stop any further pollination that could happen from insects traveling.
If something didn't make sense, let me know,
Remy

Yes, Remy that does make sense! Thank you for explaining it to me. I'll need to do some checking on E. atrorubens though, it might also be a subspecies?
-B

I did some checking around and I found one site saying this:
Echinacea atrorubens - Very much like E. pallida with purple or yellowish? Flowers. Plants are found in Prairies and dry open woodlands in Mo to Arkansas and Oklahoma and Texas. Plants bloom in May-June. This plant was lumped in with E. paradoxa with the yellow forms being E. atrorubens var. paradoxa and the purple colored forms from the west being called E. atrorubens var. atrorubens.

So I checked some more and it looks like E. paradoxa is most closely related to E. atrorubens. Doesn't it figure!
I also found a site stating there were 4 species of echinacea and 6 subspecies. I didn't find a good list or chart of how they are all split up though. I'm sure there's one out there, but I'm done looking around, lol.
Remy