I can't speak for Magic Man, but while Yaw works fine without Target, each pedal also has an axis, and it would be nice to be able to use those for lateral thrust (eg depress right pedal for lateral right thrust, left pedal for left thrust). You can't do that in Elite because the lateral thrust assume both directions are one axis, joystick-style, or buttons that are either full on or full off. Target might be useful for getting the most out of the extra pedal features, but yeah, not needed for yaw (other than response curves).
Out of curiosity, what is it about the sensors that Slaw uses that means you would never support them?
Well.... both Milan of MFG Crosswind and I independently chose the same Melexis Hall sensors for a very good reason (Milan & I actually talked about this) -- it's integrated with a small CPU with DSP module on it. It means, Melexis, the manufacturer of the sensor processes its own signal, including amplification, DSP filtering, scaling, and even curving and then output to DAC, all in one package. It means, I don't have to process "THEIR" signal. After all, "they" know the characteristics of their own signal better than I do. Man, don't make me process your signal, and an analog signal at that! It's a $5 chip (in bulk, but could go up to $11 if you buy just one). I pay $5 for an integrated solution so all I have to do is program the chip and read the resulting values. That's it. And the Melexis Hall chips even has an Automatic Gain Controller on die, meaning I can be real sloppy at magnet choices/placement/alignment. I just throw in a toy grade Neo-Magnet on it, and it worked like a charm at distance of 1mm, to 2cm... whatever. Gives me a huge flexibility in designing the mechanism for holding the magnets and the sensor chip.
With Slaw's MaRS sensor... catchy name... Slaw went with a Hall chip with output of raw analog signal, and then throw in an additional DSP processor to process that signal. It's a heroic engineering effort, no less, but ultimately can be solved with a $5 chip with minimal fuzz (and if memory serves, BMW seems to agree with me on this one, as they used an MLX Hall chip on the swivel headlamp of their cars). If all that heroic engineering effort does resulted in worthy or essential improvements on the resolution and latency needed for a rudder or joystick, ya sure... but the fact is that I don't see any of the benefit. Sure, the 120MHz Atmel ARM MCUs I use in Hempstick do have DSP instructions and they are fast (and I just got my hand on a board with new an Atmel MCU that runs over 500MHz), but I would have to put in the same amount of engineering efforts as Slaw has done for his DSP processor (maybe even more), but I can avoid all that by using a $5 sensor chip.
I mean, what do I get out of throwing in such engineering efforts? Nada! Not in terms of functionality, not in terms of dollar & cents! No way, Jose!
Note that, this is not to diss Slaw's result or functionality. It's just that it makes absolutely no sense for "me" to support it.