I think this is exactly right. Speedrunners (and film editors) can perform actions with frame-perfect precision when they can plan those actions ahead of time.

I think this is exactly right. Speedrunners (and film editors) can perform actions with frame-perfect precision when they can plan those actions ahead of time. It might be similar to how musicians can perform notes with extraordinary temporal precision, as long as they know ahead of time what the timings will be (and in some cases as long as they have external references to resynchronize with). I don't think there are any speedrunners who can be expected to give a frame-perfect response to an unexpected event. The observational astronomy version could be that a human being could synchronize two clocks to less than 0.1 second difference (by starting one when the other reached a specified time, maybe), but still not measure an uncontrolled natural event to the same precision. I'm not sure how this works with timing human athletic performance using a stopwatch, but I think I would actually be wary of trusting a stopwatch-measured time much below 0.1 second precision, even if the stopwatch displays more decimal places. I think athletic records in the timeframe this book is talking about were reported to the nearest 0.1 second and not below, and having more decimal places in our measurements has required video replay and, later, other electronic timing methods.