well before conjectures are made here it should be clear that at this time NOBODY knows the answer to your question. It is being studied but results are only slowly accumulatiing. Only this year was a pretty conclusive study on this issue publsihed, but this related to a tortoise from an arid region and not to semi aquatic turtles. the meachanisms are THOUGHT to be similar. But right now nobody is sure.
So with that said there is probably more you can gain from everybody's "anecdotal" that is, "not funded by some scientific establishment" information. what I have seen is the following
(1) those water turtles that are most lumpy (save for the ones that were incubated at too high a temp) tend to all be ones that were fed hevaily on turtle pellets/and or lots of meats
(2)some turtles end up lumpy even if they are fed EXACTLY the same diets as other of the same species (even sam clutch) while their cognates do not. Sometimes this seems to occur if they are kept in the same cage and one turtle just eats more than the others. I have one slider who eats anout three times as much (a guess) and gre faster than his tank mates (WAY faster) and had to be seperated so that i could feed him much less often while not risking starving the others.
(3) UVB seems not to be an issue as much as it has been made to be as i have seen lots of lumpy turtles kept in outdoor ponds and lots of smooth ones in UVB-less tanks. (one well known herpetologist who keeps many turtles himself doesn't think artificial UVB lights work for turtles at all) But i still think better safe then sorry.
(4) With ones i ahve seen, sometimes the turtles that willing eat the calcium pills, cuttlebone or spines of fish (as opposed to those that ignore them) seem to develop a lot better but i don't think there is enough data, even anecdotal to validate this statement
(5) there ARE lumpy turtles in the wild (just like there are pyramidied wild tortoises) but it has always been assumed that these were living under somehow sub optimal conditions.
Again all these things have not been well enough studied.
over the years i have found its better to feed less, and to constantly try to widen the variety of foods. But i have far from a "perfect" record as i have several lumpy turtles myself.