True, there probably have been some important refinements, but part of the issue is that a lot of that work has been done along very long, sometimes many-decade

True, there probably have been some important refinements, but part of the issue is that a lot of that work has been done along very long, sometimes many-decades long, paths that so far have had zero experimental validation. Basically, people have been building giant ladders and constantly improving them, but we don't know if any of the ladders were put up against the right wall. We have no idea if these theories actually correspond to reality or not. For example, string theory and supersymmetry. The LHC was created partly with the goal and expectation of detecting particles that would validate supersymmetry, but so far it hasn't, and many are beginning to believe the theory may be flawed or wrong. In some sense that is a bit of a theoretical advancement, but so far, supersymmetry theorists don't appear to have a popular alternative they're considering instead. If they fail to find evidence for it in the next 10 or so years, it's probably just a total dead end. And string theory has been at it for even longer.