Now that I realize what the problem is, any chance of adding an option to include those Matrix profile tags to fall back on like other programs do?
Possibly. I'll have to take a look at how difficult that is.
As a somewhat related question, is there any reason why Argyll refuses to make an ICC 2.4.0 Matrix profile (with A2B & B2A tables) like you can with LUT profiles?
I'm not sure what you mean. By definition A2B and B2A tables are cLUT profiles, not matrix profiles. There can be only one "matrix profile" matrix, and it is always colorimetric intent, and is easily invertible to provide the PCS -> device conversion. Graeme Gill.