There's always tradeoffs...
Originally Posted by Rafal Lukawiecki
Flash is closer to daylight in terms of spectral distribution, though it is "discontinuous" meaning it has spikes of certain wavelengths instead of a smooth curve. That brochure from IC-Racer shows the zig-zag curve and describes how they dealt with that issue.
At the low settings, your flash duration might be very short, less than 1/1000th second - so your results from too short a flash might be affected by Reciprocity Failure. (Recommend brighter flash and find a way to block some light).
Tungsten doesn't have as much blue light as flash - an 80B blue filter can compensate.
For tungsten, I think you can easily rig a reliable shutter for repeatable times (The best time is between 1 second and 1/1000 second to avoid Reciprocity Failure). Think about jerry-rigging a camera body to the enlarger lens mount.
As IC-Racer says - all you really need speed testing for is to see how one film compares to another. Consistent exposure will give you a relative speed test - where the results give you an EI to believe. When you get a value that is within 2/3 of a stop of what you expected... That is good enough. We can write pages of threads trying to find that last 2/3 stop - You would have to follow every ISO/ASA specified procedure to get absolutely certain film speed results (including taping film to a plate of glass inside a Thermos bottle).
My favorite short-cut is to develop to the "ASA Triangle" and then mark that down as if it corresponds to box speed. My shortcut is within 2/3 stop of reality.