I've noticed that the calculated luminosities and radii of many procedurally generated B-type stars simply don't make sense. Using the well-known Stefan-Boltzmann equation, I always seem to end up with radically different values from what SE displays. For instance, a B0 V star with a radius of 6.36 solar radii and temperature of 29220 K will have a luminosity of about ~26900 Sols, yet SE calculates its luminosity to be a measly 539 sols, which is a huge deviation from the value suggested by the equation. Lots of other examples in-game, but the same core problem persists.
As far as I can tell, this issue is mainly with B-type main sequence stars. Their luminosities often don't align with their temperatures and radii and are considerably lower than they should be, which leads me to think the procedural generation algorithm is not applying the Stefan-Boltzmann law correctly.