Not sure this is still a concern, but...
I can't speak to why LC gave up on the UX-129. There weren't many of these produced and I've never encountered anyone that has one (tho, that isn't saying much). The UX-129 design featured several "quirks" that were unique in comparison with the other IC-900/901 modules. The PLL chip was much more complicated and the LO featured an RIT/XIT system that used GPIO signals from the PLL chip to provide +/-7 RIT/XIT steps. See http://www.rollanet.org/~joeh/projects/ICOM_Design_Guide.pdf
for an in-depth discussion of the control protocols for the different modules (note: while I *think* I understand the UX-129, I've never actually had my hands on one). The UX-129 uses the same transfer protocols as the other modules, but there are just more transfers needed. The PLL chip has 4 registers which all need to be initialized and if the RIT feature is desired, there are a series of register transfers that are needed to control the RIT circuit. This basically consists of toggling the PLL GPIO signals which connect to an up/down counter which drives a simple R-2R D/A converter that tweaks the VCO. ACC's approach was to ignore the RIT feature, but I've never seen their controllers close up.
The IC-900 service manual speaks to some of these features, but scanning data on an IC-901 was the real trick for me to figure out how the UX-129 was controlled.