Disclaimer: I know absolutely nothing specifically about the Sony VPH-G90U, but since I already had the service manual open for another thread I thought I'd take a crack at this issue too.
If the YA and BA Boards were already found to work correctly in another chassis, doesn't that more or less leave only the MA Board?... Which almost narrows it down to basically a bad connector, solder joint, or trace.
The obvious suggestion would be to check around connector CN602 and CN603 (that the YA Board plugs in to) on the MA Board. Especially pins 9-10 and 12-16 on CN603, as those are the clock and data lines for the I2C Bus and Registration DACs. Somehow, given who's asking, I'm guessing it's not going to be quite that simple though, is it?
Do you get a ~5MHz (per the Theory of Operation Manual) clock waveform at CN421, Pin C12 on the YA Board?
Do you also see that same clock waveform at CN341, Pin A13 on the BA Board?
And again, let me reiterate, I may have no idea what I'm talking about... Just throwing out a few suggestions that sounded good to me at the time.
Edit: A couple other things I realized...
I imagine you've already tried this, but if you input a composite NTSC signal on the video input, do the COLOR, HUE, and SHARPNESS controls work? The IC responsible for those adjustments (IC1600) is on a different bus than the DACs for BIAS, BRIGHTNESS, CONTRAST, etc. (IC609 and IC611).
Also, if the D-SCL (clock) waveform I mentioned above is
present on both YA and BA boards, are you seeing periodic assertions of D-LD0 at CN421, Pin B13 (YA Board) and CN341, Pin B14 (BA Board) as well?