What if the numbers are not bad -- what if riding an Uber is safer than cab/bus/walking, but there's no way to know because the stats on cab/bus/walking are not known, so people wrongly assume they are lower than Uber's?
I just don't think the _possibility_ of a backlash (for defensible numbers, no less) justifies withholding information about personal safety. It's not like we're talking about sales figures, production stats, or other trivial things.
If people misread the numbers, for the reasons we've been talking, their reaction might put them in greater danger than currently.