More than 1M fewer students are in college, the lowest numbers in 50 years

The answer depends on institution type.

Highly Selective Institutions: the admissions process is so hands-on and personal that I could believe they are able to get a good sense of each candidate without testing. E.g., I absolutely believe Harvard's admissions folks have the bandwidth to compare grades between high schools (not that they need to). And they do all sorts of stuff that gives you a better sense for the candidate than test scores (alumni interviews, essays, rec letters, delving into performance in highly competitive extra-curriculars, exceptional community service work, etc).

Non-Selective Institutions (let's say admissions >70%): Standardized tests are kind of a waste of time and money for all involved. These institutions are functionally admitting everyone who can manage to fill out the admissions paperwork and didn't systematically fail high school courses. It makes sense to make test scores optional, because there might be a few diamonds in the 30% "oof" pile who can pull off a decent SAT score to compensate for their D-average-no-honors-courses transcript. But requiring the SAT/ACT is silly when your admissions standards are extremely low.

Moderately Selective Institutions: I definitely see utility in these universities still using the SAT. But notice that this is actually a very small set of institutions. Perhaps 100-300 the US's 3K+ colleges fit in this category.

--

FWIW, my opinion on SAT/ACT is somewhere in the middle.

I think standardized tests can be an excellent instrument when some form of assessment is necessary but more nuanced assessments of merit are cost-prohibitive. So for places like competitive state flagships, I think getting rid of testing and replacing it with an admissions process that works at least as good as testing is probably more expensive than it's worth.

However, I also think the pro-testing camp is often extremely hyperbolic. Testing is just one way of assessing merit. It has all sorts of flaws. Tests are a model, and all models are wrong.

As an aside, I'm not surprised that so many pro-SAT-the-sky-is-falling folks are mathematicians. That entire field is completely fucked up when in comes to testing. Math as a discipline is bad about intellectual peacocking in general... if you think Mensa is insufferable, spend an afternoon in a math dept. Math professors are exactly the last set of people in the world I would trust to have a healthy attitude toward the ability of testing to suss out real merit. They literally talk about their prelim exams the same way frat bros talk about hazing rituals. Systematic misuse of testing is the second biggest reason that people choose to do phds in math-adjacent fields instead of math. (The biggest reason is job prospects.)

Hardware is cheap, especially if going back a gen or 2. Software is free if open source'ing it. Piracy is also cheap too. Books and resources are plentiful, and predominantly cost time to experiment and do. And you can get hold of the developers pretty easy in open source, or dev with them.

Science has costy expensive lab equipment, knowledge is less shared and more guarded for their guild, people are harder to get to.

Engineering, depending on discipline, isn't even legal to teach yourself (professional engineer). Or it requires more expensive lab equipment in the ranges of 10's of thousands (think good oscilloscope, signal generator, rf anything). It's possible but really hard.

Math is possibly just as 'easy' as CS/IT. However, it's a super elite group. They have their own in-language that's near impenetrable for outsiders. And the people who excel are not at all easy to get in contact with.

Outside of STEM, there's other possible avenues. But given that Marx said that a capitalist society trends towards money-making activities in mathematics, science, and technology (STEM), the other avenues are probabilistically a bad bet. And yea, that sucks. But we also like to have enough money to live and enjoy life.