I can't speak from experience but the Marijuana Botany guide book says that the original Ruderalis discovered in Russia had very little THC and high CBD.
From that i'm guessing people have tried to breed in potency but it wasn't there to being with so any cross with it now is inherently susceptible to lower THC.
Of course breeders try to increase potency in new strains; that's one of the main reasons to breed in the first place.
All of the modern-super strains were created by selective breeding over many many generations from lower-potency strains.
There is absolutely no reason why this same process can't be done with ruderalis-based autos to greatly improve the potency of the offspring, and in fact this HAS been done. You just have to selectively breed the autos with more conventional potent strains over many generations, selecting each time for potency and retention of the autoflower trait. Do it enough times, and you might have something phenotypically similar to the original potent parent, but autoflowering. And again, in fact there are any number of strains like this on the market right now.
At this time the best autos still aren't as good potency-wise as the best photoperiod strains, but they are still pretty darn good and miles ahead of the original "lowryder" strains from 10 years ago.
Now, some people believe that lots of "dark" time is necessary for maximum cannabinoid quality/production and so maybe that means that you can never really do as well with 20-4 strains as you can with 12-12 strains. But again, that doesn't mean you can't still get something auto that's really good and better than a lot of conventional strains.
.