Years ago ago now, I was talking to Jerry Branch who used to be the THE head guy and his answer to whatever it was I was on about was "I don't care if you're Jesus Christ himself, without air you aren't going to make HP"....and that's stuck with me over the years.
To your point there are lot of ways to screw stuff up and make low HP even when you have enough air but my rule is get the air first because that's the hard one.
Efficiency and fuel consumption are other questions and effected by lots and lots of things......but you still need air
OK, didn't mean to touch a nerve there. however air is only one part of the equation, the engine needs fuel and the fuel should ideally be mixed properly, if the port velocity is too high then the fuel will be wetting the walls, if the walls are polished then the air velocity will slow down as frictional forces take over, basic physics at play there. they all work together in concert, to just look at air flow as THE discerning factor is a ham fisted way at trying to make more power. I prefer to look as mass air flow over pure volume velocity, for that of course temperature plays the role and velocity and pressure effect the mass flow.
lets get real technical here, where is the direct correlation of static CFM to HP? engines do not run in a static mode, you know that, it's highly dynamic with the pressure pulse and wave dynamics playing the critical role on cylinder filling. Given those parameters the engineer is going to look for ways to manipulate and use that to an advantage, boundary layers in air flow are critical, air moving at those velocities and pressures will exhibit conservation of momentum forces esp in turns, frictional forces from polished walls will slow the velocity and you will loose CFM, reduce the drag and it will increase the CFM, and that is with simply changing the wall texture. now toss in fuel with varying viscosity and temperature zones, accounting for that and designing the port so that the fuel stays in suspension is key to efficient burn and transfer of energy. the port needs to be designed for the fuel delivery system.
I enjoy complex systems and the cascade effect, bit of a masochist I think or I wouldn't also spend my time in QED research. there is no simple answer, for every choice that's made there is compromise elsewhere in the system. DI is more or less an attempt and getting around a lot of port design issues with fuel air mixtures.