Well I have never really analyzed the ins and outs of comparing the two, but I wouldn't say one was "way leaner" just because ultimately the fuel to air ratio needs to be always pretty close to the same thing regardless. The difference is in what the engine can consume at any given throttle opening and rpm, and achieving THAT is often a bit tricky. the mod does indeed (as far as I know) mimic the piston port( R5-DS7) carbs in that the needle jet/emulsion tubes and needles are near matches to the oem parts. You need the bigger air jet to go with the different style emulsion tube - not just to add more air from a tuning perspective.
Indeed, all the fuel jets are submerged in fuel. This is not actually required as a pipe could be attached that IS submerged in fuel, it is just convenient to have the interchangeable jets be at the end of the "straw" that gets the fuel. The jets are just a restrictor in the end of that straw. The fuel level inside the straw (emulsion tube) is the same as that in the bowl (well, for the sake of discussion anyway), so the higher the fuel is in the bowl, the higher it is in the straw, and the easier it is for whatever vacuum is available to drag fuel into the venturi (main bore). You can also see that higher fuel levels will cover more holes in the emulsion tubes reducing the air that percolates in. Both things tend to richen the mixture until you are running at the max flow of the jet, at which time it makes much less of a difference. Low fuel level of course will do the opposite and tend to make a leaner mix. Generally, this is not all that profound an effect especially on your thirsty two stroke. Fuel level in the bowls seems to have a more profound effect on four stroke engines, at least in my experience.
The piston port motors are rather different than your RD - more so than you might think considering how similar they are. The reed valves make a big difference in what is happening to the poor carburetors. Both rely (as do all engines) on the effect of sonic waves traveling through the air (and of course fuel) as it goes through the intake tract. Sonic waves are in the main unaffected by the material flow, so they travel backwards through the intake tract just as carelessly as they do toward the engine. (the exhaust works the same way). The waves occur because of sharp change in pressure that happens when the piston mechanically opens and closes the passageway leading to the combustion chamber. When the piston "opens the door" suddenly (or closes it), a shock wave is created that travels down the passageway - in this case out the intake tract through the carburetor. When it reaches the end- at the carb bell mouth - it sees another sudden change and another wave is created that travels back toward the engine. These pressure waves conspire with the gas (air) in the system and its mechanical motion through the passageway to either increase the pressure or reduce the pressure. Much additional power can be had by figuring out how when these events will occur and causing their effect to occur at an opportune time - say 8000 rpm for example, by adjusting the diameter and length of the passageway so you add pressure when the intake port opens for example. The extra pressure will add substantially to the charge in the combustion chamber resulting in a corresponding increase in power. The problem is that the timing of events that work great at 8 grand usually work like shit everywhere else In combination with the lack of the reeds one way valve action, the piston port engines can suffer a lot of "recursion", where air that has already passed through the carburetor and sucked up fuel is actually pushed backward through the carb, picking up more fuel, before being sucked back for a second (third?!) time toward the engine picking up even MORE fuel! In my example, the recursion would (presumably designed to) be non existent at 8K so carb tuning can be perfect there - meaning it might be horrifically bad at other rpm. Reed valve don't eliminate this phenomenon, but the mechanical motion of the air stream is substantially better controlled. All this means is that there is nothing magical about the Dave f mod. It works great on some engine combinations because all the complex nuances of that particular combination result in that carb combo working well. Some do and some do not. Always worth some experimentation though!
Recursion can be very pronounced. I had a drag bike based on an R5 motor back when that was still a fairly new bike. Very fast, but exceptionally unusable for anything else. I remember it was quite hard to start because it wouldn't run hardly at all below 2500 rpm or so. At idle (3000 rpm) you could plainly see a dense fog of fuel standing suspended in the air right in front of the 36 mm carb intake bells. If you stuck your fingers in there they came out dripping like you stuck them in a cup of liquid gasoline!
Yes tuning the mains. Getting them right can not be over stressed. NEVER adjust them for drivability. Wide open throttle at max power only. Drivability get fixed by adjusting everything else. Reducing the mains "because it drives so much better" is an invitation for disaster next time you open her up.