The reason the really nice looking jurs have thicker bodies is because they're adults. The nice ones are always thick, because the nice coloration doesn't develop until the fish reaches it's adult size. The fish we see in the hobby typically die or are sold long before they ever get large enough to display vivid colors.
I'll admit that there may be something to the regional variation hypothesis, though I find it unlikely. When one sees juruense in a fish market, all three forms of the pattern tend to be present; there are thin yellow striped mostly black fish laying next to thin black striped mostly yellow fish and mixed in with the tiger striped "cf. Rio Nanay" type. While one location or another may have a predilection for some pattern over the other, the only anecdotal data available indicates that all forms can be found along side of one another.
Correlation doesn't imply causation, however. We in the hobby must use the data that's available when we diagnose our fish. For example, I without question know that some African tiger fish that we call tanzaniae are in fact an undescribed species, but I lack sufficient collection data to properly describe and name this haplotype, so I leave it alone because the fish fits within the overall published scientific data for H. tanzaniae.
As hobbyists, we're stuck using published data. I don't oppose people who want to call nice jurs flash zebra or something. However, until Dr. Lundberg or some other equally esteemed icthyologist figures out where the speciation lies, than what constitutes a flash zebra vs. a normal jur is purely and unequivocally in the eye of the beholder, because both fish fit the published description of Brachyplatystoma juruense. Keep in mind that I say this as someone who earns his living as a fisheries biologist.
It's also worth noting that they appear to be about half the size.
I think that there are three color forms with no really distinct differentiation; it's a spectrum of the possible. I don't think that they constitute different species.
