Well, I don't think you can make any case for Variable #1 and Variable #2 being less vague, but I'm pretty neutral on that point. Variable #2 does relate to the y axis surprisingly frequently, but not fulltime. Ideally, though, you'd see JCS-like behavior where the five variable bytes rename themselves accordingly depending on the modifier and behavior. That would take more work, and you'd need occasional double-duty for cases when both modifier and behavior referenced the same byte, but it would be a lot more user-friendly.