Obviously we cannot indefinitely build larger accelerators, and if it turns out that there is no support to go at least one size bigger, then perhaps our best hope is to find more effective ways to use the existing ones we have.
As a random example to illustrate what I mean: https://www.google.nl/amp/s/phys.org/news/2018-07-plasma-siz...
Maybe stopping here will force more thinking along these lines? I hope so, anyway.
Basically if one idea out of a hundred major ones really takes off, that's considered a success for a researcher. We're talking one, maybe two real hits throughout the entire scientific career. Researchers are well aware that much of what they do and publish is of little consequence. And they have to make "empty promises" from time to time to continue getting funding from the easily distracted politicians and upper management. They also almost have to have lofty, aspirational goals to keep the motivation.
Basically 99% percent of work of a good, hard working researcher consists of going down the alleys to see if they are blind, which they nearly always are. But the remaining 1% often makes the whole thing worthwhile to all parties involved. The question is, can you pull off that 1% hail Mary or not.
The potential discoveries we're talking about in this case could upend industries, reshape the world economy, and bring about the post-scarcity future, at least when it comes to energy. If we're willing to strain credulity a little, they might even enable interstellar travel. Is that worth a few days of the Pentagon budget? To me, it is, as long as we maximize the chance of the aforementioned Hail Mary.
In contrast, an engineer that doesn't succeed with her ideas at least 50% of the time would quickly find herself out of a job, so by definition the amount of risk engineers take on is as little as possible.
This is a fair debate to have, even if my personal preferred answer is “do both”.
A typical experiment is to accelerate a lot protons and a lot of antiprotons (or other particle), make them collide, and see the "debris".
The "debris" can be a lot of things. So you must filter some of them. For example pick only the one that produce exactly two photons and two "jets" of mixed particles that are difficult to identify. Now you calculate the energy of the photons and make a nice graph, of number of collisions versus energy of the photons. Using the standard theory you expect a smooth curve, but if there is a new particle you will see a peak somewhere.
If you see something, you only see the peak. So good luck imagining what the particle can be and how it interact with the other particles.
Another team peek another filter, like the experiment that produce two photons and electron and a positron. They make a nice graphic and hope to see an unexpected peak.
Another team ...
If the current theory predicts a peak somewhere you should find it, but if you find an unexpected peak yuo may have founded a new particle.
The theoretical frameworks are useful to guess which filters may be interesting and give a chance to find a new particle. The problem is that there are just too many possible filters and you can't try all of them.
It just pisses me off that we spend trillions on bombing random people in the desert, yet a few billion for e.g. fusion research is an insurmountable obstacle.
As an aside, I think this is a great example of a role which fundamental science plays in society: It can facilitate step changes in technological progress.