The later versions of the HF Active Antenna have a 60 MHz low pass filter between the balanced FET pre-amp and the driver amplifier feeding the ethernet cable.
The low pass filter was originally added after tests with one of the first customers who lives in New York. He was line of sight with the high power FM broadcast stations transmitting from the high buildings there and he had a serious overload problem. The low pass filter prevented the driver amplifier from being overloaded so it cured his problem.
At 144 MHz the filter has about 45 db attenuation so it's possible that it's the FET pre-amp that's being overloaded. It's not easy to design a filter for the input to the FET amps as the input impedance is around 1M ohm and it would probably disturb the 180 degree phasing between the inputs.
What's the length of the whips? If each whip is 1m long then it would match perfectly as a half wave to the FET amplifier on 2m and the main lobe would fire right at your 2m antenna. It might be worth trying a longer whip say 1.5m long as that would be a 3/4 wavelength on 2m with a mismatched feed impedance of around 90 ohms and the radiation pattern on 2m would be two lobes firing up at 45 degrees and down into the ground away from your 2m antenna.
I would usually recommend mounting the antenna horizontally to add some cross-polarisation loss between the 2m antenna and the HF antenna but that would defeat the object of having an omni-directional antenna. It might be worth trying it just to see if that cured the problem and also to give an indication of how bad the overload is.
I do like your mounting system for the antenna. It's the simplest method I've seen for mounting one of those antenna in a vertical position without the feeder affecting it.