I was considering a similar line of thinking. However, there were a few hangups which I believe prevent yee-average-human from testing the glasses.
Many sites suggest a transmittance of 0.0032 (welding shade 12) to be the minimum to be safe (and typically recommend 13 or 14, which has even lower transmittance). That means you're only seeing 0.3% of the light get through the goggles. Your test would need to have enough sensitivity to pick up that tiny amount of IR light, and to not only receive a signal but be able to meaningfully measure its amplitude. You would also need to ensure that UV light is blocked as well... and all the other IR bands (I am not confident as to how wide-band protection one actually needs).
Even then, you'd have to deal with the issue of having built your own contraption to test a device which must protect your eyes from a permanent damage that you literally cannot see. I'd want to be quite the confident electrical engineer and systems test expert to make sure that the tests I do actually test what needs to be tested.
Note: I have a related question asking how one can track down the ISO certifications and see if your particular vendor has legitimate glasses or not. I run into the same issue: how can you trust that this device performs as indicated, because your eyesight depends on it.