Hi Tom,
These are my views as an ordinary AAVSO observer not as someone who could be regarded as an expert.
First, I suspect that most AAVSO observers who transform their results do not correct for airmass. Second, I believe from various readings over some years (including comments on this Forum) that airmass corrections can be dispensed with if (1) the range of airmass at which the observations are made is relatively small (for example, X < 1.5 for all images) and (2) the field of view is small (say, 30 arcmin, or maybe less than 60 arcmin). With respect to the field of view, what really matters is of course the distance across that part of the field occupied by stars measured. I don’t use these numbers to mean that they are definitive values - they are just examples.
That said, I presume that if an observer wants the best results possible correction for atmospheric extinction is the way to go. Can it be done? Of course. What is the short procedure? I’m not sure there is one.
As I understand it, if transformation coefficients are unknown, the extinction coefficients can be determined on a clear night by following an A0 star across a range of airmasses, and plotting instrumental magnitude for each filter against airmass. The slopes are the coefficients, for example k’v and k’b.
If the transformation coefficients are known the procedure, regarded as ‘simplified’, is to observe standard stars of any colour at a range of airmasses. Instrumental magnitudes are calculated. Plots (for B and V filters) are drawn of (V-v)-Epsilon(B-V) against airmass, and (B-V)-Mu(b-v) against airmass. These slopes will yield k’v and k’bv respectively.
Second-order extinction coefficients are determined by following a closely spaced pair of red-blue stars across a range of airmasses, and plotting delta(b-v) against X*delta(b-v) where X is airmass. The slope is the second order extinction coefficient.
Once extinction coefficients are available instrumental magnitudes of variable, comp and check stars are corrected for airmass. (Corrected to what has always interested me. Airmass of zero, correction to extra-atmospheric? Airmass of 1, correction to values at the zenith? Correction to an airmass of zero involves a substantial extrapolation. Correction to an airmass of 1 is a more limited extrapolation).
Instrumental magnitudes corrected for airmass are then entered into the transformation equations for the determination of transformed magnitudes.
The above is my understanding from intermittent readings over many years, and a small amount of experimentation with airmass correction. I do not routinely correct for airmass. I do transform my results when I believe I need to (e.g., always for DSLR photometry), but dispense with transformation for long time series where the objective is timing of maxima or minima and comp stars close to the colour of the variable can be selected (edit) and observations are through a V filter.
Most of my observations are at an airmass of 1.5 or less because I have compromised horizons.
At normal air masses up to say 3 maximum and good transparencies the error contribution is very minor. More hand waving. 
