So, what you're saying is that in order to keep 60FPS consistantly you need to cram the entire rest of the frame into 6.7ms? (That's an equivalent of running at 150FPS "normally".)

People forget how sensitive framerate is to seemingly minor amounts of time.

(Even if you have a target of 30FPS (please don't!), that's still an effective target of 43FPS.)

Nursery gc is proportional to the size of the live object set (here alleged to be 0) plus a constant overhead. It's not a simple thing to benchmark.