I didn't read the whole paper you linked, but doesn't this sentence in the abstract contradict your last sentence?
>Results indicate that camel casing leads to higher accuracy among all subjects regardless of training, and those trained in camel casing are able to recognize identifiers in the camel case style faster than identifiers in the underscore style.
yes it appears that way without reading the paper. Basically the main factor to me is speed which has a huge differential for untrained subjects "The model finds that identifiers written in the camel case style took 0.42 seconds more time." Doing the math you see how much wasted time that is. So I should have been more specific in the parameters. Your quote is perfectly valid as well.
>Results indicate that camel casing leads to higher accuracy among all subjects regardless of training, and those trained in camel casing are able to recognize identifiers in the camel case style faster than identifiers in the underscore style.