PuffyCloud It sounds like you measured the "FfT" bulbs you tried. Did you measure any frequencies, and what exactly do you mean by "they go dark"? Like, really no light anymore? This might be very important, please reply.
I just measured my headache bulb which I still own with my oscilloscope setup. It is the "GreenAndCo TM-A19-6W-E27-W". I did not find frequencies other than 50 Hz and 100 Hz apart from very small fluctuations around 60 kHz. The latter are so small they only truly appear in the spectrum analysis (and only in AC mode). And the visible 100 Hz ripple appears to be very minor, too. At most 5 mV out of 600 mV total amplitude. That is well below 1 % difference. Can such small differences really cause headaches?