The plan is MAF because I don't have a clue about what to expect with a speed density system and if all goes well this will be a miller cycle engine with a whipple supercharger in a year or two and the drawbacks of having an MAF in the induction system go away when you are stuffing it with a positive displacement supercharger.
The reality of it is, it can be programmed either way, I just don't like to use an implicit solution that requires monitoring temps, pressures, etc. where if your sensor drifts with temperature or you change something about the induction system and you are immediately tasked with recalibrating and re-characterizing the sensor. MAF is simple and I understand it. I'll find a sensor that can handle 110% of what I expect to ever need and all tuning can be accomplished in real time from a terminal with the engine running.
