First of all you are incorrect as to Hilbert. he ws NOT a fan as far as experimental physics. I have read his correspondence, and he did try to largely axiomize physics. So you are factually incorrect
Secondly, I am aware of the massive Graviton scheme you mentioned and I used "enhanced quantization" as a way to do the problem
See Klauder here
2015 Klauder JR. Enhanced quantum procedures that resolve difficult problems Reviews in Mathematical Physics. 27. DOI: 10.1142/S0129055X15300022
I disagree with the Bimetric Gravity approach.
And I think you missed the main point of my essay. Which is that certain inputs have to be experimental. Here is a story I can share which had to do with what is called the Clausius - Clayperon model of Dark Matter and Dark energy
It has, roughly density = - constant/ [(pressure)^alpha]
Starobinski whom I have met repeatedly in Marcel Grossman 13 and 14 in about 2000 or so did research with the Clausius - Claperyon relationships in early universe conditions and matched data sets with an alpha = .857 or so.
In String theory, if alpha = 1, it is in fidelity with respect to that theory.
Starobinski had a different value which matched experimental conditions, but was off from the String theory mandated alpha = 1
What you missed is that although our models are in certain cases useful that we do NOT have the ability to avoid experimental inputs, and that this one about the Clausius - Claperyon relations as to a joint Dark Matter - Dark Energy model is a classic mis match between string theory predictions and data sets.
If the alpha were = 1 in early universe conditions, the early universe in terms of Dark Matter and Dark energy would be very different
See Introduction to Cosmology 3rd Edition
by Matts Roos