The points your making make sense. I am thinking of it like this: Say there is an elegant argument. It checks out at first. Then you do the unit analysis, and find out the units don't match! But you still don't find the flaw in the original argument; maybe because it's suble in some way. That's where I am: Very smart people have been writing things in the vein of your post here for millenia, and it always seems convincing in the light of contemporary knowledge! Then is proven to be incorrect by major advancements.

Perhaps this will help: Indeed high energy physics is a very high budget project! But there are many areas of the natural sciences which are not high energy physics. This area has been a big deal over the past few decades, and I wonder if it's an over-commitment at the expense of other areas.

You can do many molecular bio lab techniques with a budget of $10k in equipment and reagents, for example. (If used/entry level) I believe there are also many areas in science, chem, and bio which can be done on a theoretical level, or with computers, etc.

Another angle: We are in the earliest steps of neuroscience. Many biology tools and techniques are borrowing something serendipitous we found in nature (CRISPR, TAQ polymerase etc, leveraging living cells' equipment to produce proteins etc). We have no concept of a general chemistry simulator. Molecular dynamics simulations can only work on very small systems for very small timescales, and are based on many approximations, and assumptions which provincialize them. We are very likely missing a big picture of the lower levels of GR/QM. It is very hard for me to agree with "Yep we're good; nothing left to discover here without really expensive equipment!".