It's a nice story. Some statistics might help motivate concern. E.g., The disproportionate risk per mile for motorcycling is chilling. But in California "rear end" as a category appears to be only about 15% of all MC accidents, and even then the category doesn't distinguish being hit from hitting.
I suspect a lot of the effort here is marketing, so you might extract extra value (and do more good) with a complementary product.
A big benefit might come from controllable red light and warning bell for the rider when you're more than 15% over speed limit. Speed is the biggest risk by far, and also the risk you can control (since no brake light can defend against the inattentive driver). It's way too easy to not realize how fast you're going.
Indeed, for that you could also factor in stopping distances. Starting out with lighter, nimble bikes, I got in real trouble with the 650+ lb monsters that dither about stopping. Those would get a lower threshold. And you'd apply a higher threshold for open road vs. traffic, etc.
Then build out a UI workflow and logging, and it might grow into an interesting intelligent copilot product line addressing all the pilot-controllable factors. As a motorcyclist, it's really on you to save your own bacon.