A reasonable question; I can think of a few different ways to answer it.

We could look at other places with safe cycling infrastructure, and compare to their statistics (the Netherlands has around 1.4 killed per 100 million kilometers cycled, for example).

We could also answer it with surveys on the perception of safety. This has some downsides, because it is subjective. But the subjective feeling of safety is also important, because facilities that do not feel safe will not be used.

We could also look at the deaths and injuries that do occur, and consider if they are really bizarre black swan events (person on bicycle eaten by escaped zoo lion), or things we actually do know how to prevent/greatly reduce, like doorings, right hooks, etc.

Right now, we often see people die from events we know could be mitigated by better infrastructure, and that such things could be practically deployed.