It would divide by much more than that: It would turn any high frequency into 400Hz, since that is the highest frequency that can come out of it.If I follow you, that circuit will divide by 2 after the high limit. Triggering on every other pulse.
For sensor frequencies <<400Hz, the timer always resets long before the next sensor pulse arrives, so no problem. For sensor frequencies >> 400Hz, the one-shot gets retriggered almost immediately and the output is at the maximum 400Hz determined by the 555. (The output pulse is a constant interval.)
The timing gets confusing (to me) for sensor frequencies just over 400Hz. The one-shot pulse ends, and then you await a trigger from the sensor. You just missed one and have to wait for for almost one entire cycle for the next sensor pulse. But that's a small impact on the overall pace of the 555, which will be running increasingly closer to its limit.
I suppose this is not a precise approach. 100Hz and 1000Hz would give precise results, but there is a rounded shoulder as the frequency approaches the limiting rate.
It'd be nice to see a simulation.