
I find that this kind of science needs to be done more often. Maybe I am still young and head-strong, but it seems that too many people get 'stuck in their ways' and don't try to buck the current paradigm. If there are too many approximations and generally it is agreed that a better solution is needed, why not go back and question the underlying assumptions? I try to do that in my research quite regularly. Just because conventional wisdom says that X or Y are not good candidates doesn't mean than a combination, XY, can't be a good candidate for some new property. We don't have an intuitive enough sense of how properties vary in binary (Much less ternary) phases (Such as 2 or 3 elements combined into one material) to rule something out. Similarly in theoretical physics/chemistry, there are many problems with how to model and accurately reproduce experience without the calculations taking a lifetime to finish. I am glad to see someone making strides in this and coming up with a new paradigm that more accurately and succinctly describes reality.
You must be new here, in the physics works. Joke aside, don't believe that there is not progress and constant work in almost all scientific sectors. I have worked for 4 years in nanotechnology lab and I am a physicist and believe me, there are numerous challenges that have to be overcome in some problems or research. Don't forget that theoretical physics and mathematics are some of the most difficult sectors and some problems can take more than 100 years to be solved. This does not mean that they are not worked or their research has stopped, just that they are too difficult and too delicate to get the proper approach. Also, theory predates practice and, almost, always a theory can be 100 or more years before the technology manages to observe or do it.
Some examples:
* Gravitational waves observed in 2015 - theory in 1915
* LHC 1st completion in 2008, took 10 more years for the theory behind it, 10 more years for technology to catch up and 10 years to build· so the whole work started from 1970s.
* Honeycomb conjecture solved in 1999 - theory in c.290-350BC(!)
* Dodecahedral conjecture solved in 1998-1999 - theory in mid 20th century
* Fermat's last theorem (Practically) still unsolved - theory in 1637 (Although it starts in ancient Greece with the Pythagorean triple).