The method I describe appears to work at all shutter speeds. Full results for the camera I tested are in the attached image. The red lines are +/- 1/2 stop from ideal.
View attachment 389221
I use the same setup to measure speeds on cameras such as copies of early Leicas (don't own the real thing) where a beam cannot be transmitted through the film plane to a detector. In those cases I put a piece of reflective film on the pressure plate and bounce off that.