The amount of gas flowed is only important in the power output of the sound, all else being equal. You can't take a linear measurement and ascribe a volume to it anyway; you forgot to mention the bore. But that's irrelevant.
Sound is pressure oscillations, and you shouldn't confuse traveling waves for standing waves. For an engine running at 2000 rpm, the fundamental frequency you will "hear" is 17 Hz (you can't really hear that very well, so you hear the harmonic series above that which "suggests", psychoacoustically, the fundamental instead; you can probably feel it quite well, though, which is most of the fun). That is true no matter the bore or stroke dimensions; a two stroke ("2t") motor, however, will have its fundamental at 33 Hz instead.
Because a microphone is picking up traveling waves coming from the exhaust tip, the placement only affects the phase, not the amplitude, of a given wave. Except, of course, where there is phase interference of multiple waves. For phase interference to cancel entirely (anti-phase: 180 degrees), the distance from different parts of the source to the microphone has to be half the wavelength. For that reason, wavelengths that are close to or smaller than twice the size of the source are more affected by such interference patterns. For a 2" exhaust, we're looking at frequencies approaching 6750 Hz and higher. Anything lower is barely affected at all. The wavelength at 17 Hz is 20 metres, such that moving the mic by 1 metre gives a phase difference in the traveling wave of 18 degrees (effectively only a time delay), but the maximum phase difference available for interference from the 2" source is only 0.5 degrees (less than 0.01% amplitude reduction, from the cosine). At 6750 Hz, moving the mic only 4" can move from one node (point of maximum interference) to the next.
So mic-placement is critical for higher frequencies, not bass frequencies. After a distance of about 6 times the source size (from memory), the only concern is that you record on-axis, as the interference pattern simplifies greatly to a large central "lobe", with "side lobes" at different angles for different frequencies. In practice, to avoid wind noise, you record just off-axis, and perhaps double the minimum distance (mostly for SPL reasons, I'd guess). So two feet is perfectly far enough away for good fidelity from a 2" exhaust. Speaker cones are pretty good approximations (see
here): the lack of an acoustic baffle on the exhaust pipe and the mean flow through it complicate the analysis, but the measured pattern is broadly the same, only perturbed from the "ideal case", in terms of "lobe widths" (angular distances between minima for a given frequency) etc. PD used such a model for their sources starting with GT5P, which inspired me to do the same.
The turbulent noise is a separate thing, and has to be handled separately in the game. It's hard to separate it from the pressure trace of the exhaust sound in a recording, though. That turbulence modulates the exhaust sound, much as tyre noise and the aerodynamic wake of a high-speed car gives the "fast jet" sound, to give a washy effect: iRacing and pCARS use recordings from distances where this becomes very apparent.
I suspect the idea that "bass needs distance to evolve" is due to a misinterpretation of the "preferential" attenuation of higher frequencies over distance. So it's not that bass increases, just that treble decreases; except in the case of the above-mentioned turbulent modulation (converting low-frequency energy to high frequencies, filling in some of the natural atmospheric attenuation).
Note that this is a completely separate issue to the whiny engine (as opposed to exhaust) sounds in the game. Mechanical sounds are apt to be biased to higher frequencies, especially when the mechanics involved are high frequency themselves. The lack of intake sounds (usually very bassy) really hurts there.
. 
