Ninety-three percent of RTOR pedestrian or bicyclist crashes resulted in injury.

So, one of the data sources they use is for fatal injuries only and it appears that right turn on red accidents are not usually fatal. Ok, but look at that injury rate; injuries that are not fatal but could still be life-changing.

That article also talks about the limitations of the second data source they use

My overall reaction to that article is not “meh, no big deal”, but “crap, we should have better data on this”. Anecdotally, I’ve seen much worse driving behavior since COViD, where it’s becoming all too common for cars to not even slow down for right on red, and people here online are trying to defend that you don’t even need to stop despite that being clearly stated in the law. I do have a nice walkable downtown, but walking it has been getting more dangerous in recent years: if you hit my kid because you didn’t feel like stopping, it won’t be at all comforting for you to say “meh, it’s not a fatality”