[Luce]

My understanding is the hall puts out enough of a signal to directly drive the optoisolator without the VR conditioner circuit. The VR conditioner greatly amplifies the signal, and the induced noise as well, to the point that MS thinks the noise is a trigger. But it's where MS knows there shouldn't be a signal, so it loses sync. That's why when the engine was sitting still, you were getting little blips of tach readings.

I don't know where the idea to run the hall through the VR circuits came from other than someone realized it would work (sometimes) and made it easier to market the MS as a one size fits all kit. I see no other reason for it. Since you opted for the assembled and tested ecu, I'm assuming you're not comfortable with changing the jumpers, but that would be my first attempt at a fix. If you dig into this, make sure the optoisolator, U3 is present as well as D1, D2, R12 R13, C12 and C30 are present. Also make sure you have all 3 jumpers in place.
To try this approach-
TACHSELECT needs to be disconnected from VRIN and connected to OPTOIN.
XG1 and XG2 need to be connected.
TSELECT needs to be disconnected from VROUT and connected to OPTOOUT.
Be sure you can locate all 3 of these sets of jumpers and that you do have a quality low wattage soldering iron and the skills to use it. This is a fairly densely populated multi layered board, and you can ruin the whole thing.

If you still have noise issues, disconnect XG1 and XG2 and use a vacant DB37 pin and run a wire from XG1 to that vacant connector pin in MS box, and run the wire to the ground on the hall sensor. If possible twist the hall signal and ground wire together. This eliminates most of the induced noise.

In my opinion (again, take it for what it's worth...the musings of just another internet troll) using the VR circuits for anything other than a VR pickup is like hooking the speaker output on a home stereo to the microphone input on a tape deck to copy your buddies albums to tape. Yes, if you turn the stereo volume down to the absolute minimum while recording, you may be able to make out what song you're hearing, but it will be far short of enjoyable music.

If someone can present an argument for the merits of using the VR circuit with a hall sensor, I would love to hear it.