I don't usually us M-S because of the inaccurate low freq reproduction...
When an omni is used as M mic the LF is much better with M/S.
The S-channel benefits from using a real figure of eight, not a double-cardiod with right side phase reversed.
The figure of eight will deliver only few LF due to it's natural drop in the frequency response, so the omni dominates here.
One thing people don't realize when comparing directional with omni mics:
The omni records sound pressure (air pressure), the directional mics pressure gradient (air movement).
If you look at a soundwave, the maximum pressure is where the minimum movement is.
This is specially important for standing waves like room resonances with wavelenght of several meters size.
It's one reason why those mics sound different in the LF range, and they might need different placements for sounding good, besides the direct/reverb balance issue.
When matrixing the M and S channels, the S channel that is in phase with the M channel shows about 4dB higher than the reversed S channel.
The stereo field sounds good when the in phase side is lowered by 4dB.
I guess your USM 69 does not deliver the same level for both sides of the capsule that was used for the S-channel.
4dB difference would mean it's closer to hypercardiod then figure of eight.
If you still have the mic you can check that.
Something similar happened to me long time ago when I used a pair of KM56 intended to work as crossed cardiods.
One of the mic's switches delivered a problem, the mic switched itself into omni.
I could fix that later by turning a mixer into a matrix.
By carefully adjusting the inphase and phase levels for each channel and side I could build a normal sounding stereo image.
Regards
Kai