The direct detection experiments have been running for decades now: LUX, XENON1T, PandaX. All null results. At some point you have to ask whether WIMPs are even the right model, or if MOND-style modifications to gravity deserve more serious attention than they get. The physics community has been pretty resistant to that conversation.
MOND gets more traction than you might think in certain corners. Stacy McGaugh's group has been pushing this for years. The problem is that MOND handles galaxy rotation curves okay but falls apart at cluster scales. Dark matter handles both, which is why it's still the dominant framework. That said, your broader point stands: the null detection results are genuinely difficult to interpret. Either we're looking in the wrong energy range, the coupling is weaker than predicted, or the particle hypothesis is wrong. None of those options are great news for the standard model.
This is the thing that keeps me up at night. 27% of the universe is something we can't directly observe, measure, or touch. We just know it has to be there because nothing else makes the math work. That's wild to me.