It is indeed unpractical to keep up with all variations of LORETA and recurrent claims of increased accuracy, despite the original sLORETA supposed to produce zero-error. The approach certainly has merits and we would be happy to hear directly from your own experience so that we may re-prioritize a possible Brainstorm integration.
I think you are right, Loreta has a lot of variations and everyone claims to be better than the last one and It won't be possible to integrate all of them. However, I think it might be interesting to give e-Loreta or z-Loreta a chance someday in BS, that would keep this software up to date
But in my experience, Mean Entropy approach has more accuracy