The inherent problem with trying to determine shutter speeds of a mechanical between-the-lens shutter with this level of precision is that this type of shutter is never 100% efficient, and not as efficient as, say, a focal plane shutter. Older photo books demonstrated this issue with successive images of a between-the-lens shutter as it opens and closes--it opens starting with a small aperture, growing to fully open, then closing down to the small aperture, then closing fully--like a round window growing open from the middle to the outside, then reversing.
Thus, the shutter is only fully open for a percentage of the exposure time, and that percentage varies with the f-stop being used, and, of course, lens diameter. At a small stop, the efficiency may be 90% (I am guessing here), and at a large stop, it may only be 60%. So trying to nail in your shutter speed to within 1/3 stop may not be practical for photographic purposes.
Where shutter speed testing can help is to find really big discrepancies--like a shutter speed that reads 1/50th of a second and is really 1/4 of a second.