Ok...well that is a lot of work on your part...thanks.
Even though we'll never really know the real production figures on any of this, because the 'K' is such a special Kar, these sorts of exercises are very worthwhile.
It of course demonstrates that any Kar is a rare beast and worth the amount of effort that people spend on researching (through this site in particular), owning, restoring and maintaining.
And no matter how the numbers play out for the assumed production figures of select K models ie. fastbacks v coupes v convertibles or GT v non-GT or one color v another color or one option v another option etc, the numbers, it must be accepted, are incredibly small.
I've done a similar 'statistical extrapolation' exercise for a 66 K GT Auto Coupe that I own, just based on Gregory's book registry (which is a sufficiently large enough representative sample to draw educated inferences). Therefore, being biased towards my Coupe, I'm glad to find agreement that around 50% of all K-Codes were fastbacks (2+2). Further, I estimated around 33% would have been Coupes, and the remaining 18% or so Convertibles. So, in my world, I've got a rarer (although not necessarily more desirable) Kar than a fastback.
So I'd estimate of the 5,469 K-Codes produced in 1966, only about 1800 (33%), would have been Coupes. Moreover, how many of those would have been GTs? (I'll defer to your calculations above), but not many.
Furthermore, from a further study of Gregory's data, only about 25% of K-Codes would have been automatic (obviously affected by not being available in 1965), so the figures go down again.
But, this level of speculation of my part can be trite, and I'm just happy enough I own a vehicle that not only looks and drives great, but has a undeniable place in Automotive history because of, amongst other things, its performance characteristics, racing pedigree and rarity.... cheers.