Okay, with everything there are assumptions. Tables and computers are the same.
So, i ask you a question, just one question:
How do you imagine your body??? is your body:
a) A slab, a solid slab of everything
b) A solid pool of fluid, where only the fluid dynamics need to be considered
c) A composition of a 8, 12, 16, 20, 24, 28+ types of body tissues, some hard, some soft, that operate independantly to each other
d) A composition of a 8, 12, 16, 20, 24, 28+ types of body tissues, some hard, some soft, that interlink with each other
e) none of the above
f) a comnbination of the above
And hence we have the problem. Each computer / Algorithm, takes an assumption on the above, and then fudges it.
The VPM model in its true model, bubbles people. Hence it has been "fudged" a number of times to get the present model.
The straight VHL -16a, hurt people (what you call the Buhlmann model) so they introduced b, which is a conservative b, and then in computers they introduced c, which is a more conservative version of b
HOWEVER, the RGBM model is a VPM model, which is fudged to act like a VHL model... and thus you then get into the discussion about how much should you fudge a model... for me personally, i fudge the VHLb to a gradient factor of 70/20, or 75/15, NEVER any more on either, and thats following advice from an aclaimed hypobaric doctor.
Ohh and also, off all the documents ive read, the only one which always stands true, is Haldanes surface tension analysis, that shows that your body should include a stop at 1/2 its maximum depth, to reduce bubble tension..