I spent a year analyzing the response of light through a shutter and various sensors before I released my tester to the Kickstarter world . One thing I learned is that there is an effect called prenumbra which means that the change of light with respect to time does not "snap on".. It comes on somewhat slowly due to the light getting around the edges of the shutter curtain or vane. Looking at the output of the photodiode, even an extremely fast one shows this effect, The light wave form does not have all that fast rise and fall times, and appears to be slightly trapezoidal when viewed on the scope. Using the typical photodiode and capacitor acts as a differentiator circuit, This will tend to actually make this trapezoidal wave look more like a pair of spikes, which is what the audacity program displays. The problem with these spikes is that they don't occur at the time of full shutter opening. These spikes that are registered on audacity indicate the very beginning and very end of the shutter event, not the time when the shutter is at maximum opening. My tester does math on the multiple samples of the light sensor, and determines the maximum opening point, which is where the timing is measured.
I do realize that the simple three component shutter tester will get you close, and as "shutterfinger" commented, the difference may not be noticeable in the print. Shutter testers are most helpful to determine timing consistency and shutter capping, which can cause exposure problems if severe.
.