Point being: these researchers were chosen by, and paid by Mozilla to do this. The researchers have a strong incentive to be biased; to leave out details; to p-hack. Though they claim to be neutral, no human is capable of being unbiased, especially when his paycheck depends on it.

I have nothing but utmost respect for Mozilla and these researchers, but the paper needs to be read not as facts, but merely conjectures, to be verified by other more neutral parties.