a) Yes, QCD gives the nucleon-nucleon force law.
What is precisely the nuclear potential of the deuteron given by QCD?
Solving it has been a multi-decadal effort which is hard to summarize for a nonexpert; there's a top-down approach, (google for "Argonne V18"), which takes as inputs the particle-physics observables (i.e., nn and np scattering experiments) and gives as an output the effective two-body potential seen by nn, np, and pp pairs inside a nucleus; the bottom-up approach of chiral effective field theory (
http://arxiv.org/abs/1105.2919 for a review) which solves a series of increasingly-accurate approximations to QCD. And nuclei are multi body systems, so all two-body knowledge has to be patched into many-body physics.
"Argonne V18" is an empirically fitted potential. In his paper, Machleid gives many numerical results but none of a single binding energy. In textbooks, the deuteron binding energy is calculated from the wave function calculated from the binding energy, a good joke!
I remind you: it's incumbent on you to know this sort of thing
before you guess (or infer, or presume) that no such approach works.
My approach works as you can see in my paper "Electromagnetic Theory of the Binding Energy of the Hydrogen Isotopes" http://www.springerlink.com/content/h673n477n243vu46/export-citation/?MUD=MP
b) No, QCD calculations cannot be carried out
to high precision to give this force. If you think that is a problem: the laws of E&M and quantum physics cannot (or could not until the 1970s) predict the
color of metallic gold---are quantum mechanics and E&M therefore wrong? No, rather, it's just that d-orbitals are computationally messy. The law of gravity cannot predict (or could not until the past five years or so) the infall path of colliding, spinning black holes. Does that mean GR is wrong? No, it means that it's hard.
Of course complicated structures are difficult to calculate but the simple structures of the same type should be computable simply and, with the supercomputers the complicated structure should be calculable. Colleagues of Machleid have calculated the binding energies of the helium isotopes (how about the H isotopes?). For the N>2, assuming a zero binding energy of the excess neutrons, I get a better result than the super computer (see my horizontal line between the supercomputer calculated and experimental values on the graph). Their excuse is that they don't have the 3-body force!
[/url]
"The calculations are too hard" is different than "the theory has a problem". There is no law of Nature saying "The Universe will only contain simple 1/r^2 force laws." QCD is computationally very hard in some domains and easy in others. It has passed all experimental tests in the "computationally easy" domains. As the theorists get better, it's also increasingly being tested (at the 1% level) in difficult domains, like meson spectra.
The only proven forces in the universe are in 1/r2. For the deuteron, the theory should be simple at least in a first approximation.
c) Nobody understands the weak interaction? What? Maybe
you don't understand it, but don't project that onto
us; the weak interaction is on par with QED in being both computationally-tractable and experimentally-tested. I invite you to name
one weak-interaction-theory prediction that disagrees with experiment; I invite you to name one weak-interaction experiment for which the underlying theory is intractable.