https://onlinelibrary.wiley.com/doi/full/10.1111/ina.13070

For most of human history, the dominant paradigm was that many diseases were carried by the air [... the] miasmatic paradigm was challenged in the mid to late 19th century with the rise of germ theory, and as diseases such as cholera, puerperal fever, and malaria were found to actually transmit in other ways. Motivated by his views on the importance of contact/droplet infection, and the resistance he encountered from the remaining influence of miasma theory, prominent public health official Charles Chapin in 1910 helped initiate a successful paradigm shift, deeming airborne transmission most unlikely. This new paradigm became dominant. However, the lack of understanding of aerosols led to systematic errors in the interpretation of research evidence on transmission pathways. For the next five decades, airborne transmission was considered of negligible or minor importance for all major respiratory diseases, until a demonstration of airborne transmission of tuberculosis (which had been mistakenly thought to be transmitted by droplets) in 1962. The contact/droplet paradigm remained dominant, and only a few diseases were widely accepted as airborne before COVID-19. Airborne transmission is a major mode of transmission for this disease, and is likely to be significant for many respiratory infectious diseases.

https://royalsocietypublishing.org/doi/10.1098/rsfs.2021.0049 gives an excellent overview of the scientific consensus regarding disease transmission pathways over the last ~150 years.

[COVID] recommendations have been based on four tenets: (i) respiratory disease transmission routes can be viewed mostly in a binary manner of ‘droplets’ versus ‘aerosols’; (ii) this dichotomy depends on droplet size alone; (iii) the cut-off size between these routes of transmission is 5 µm; and (iv) there is a dichotomy in the distance at which transmission by each route is relevant. Yet, a relationship between these assertions is not supported by current scientific knowledge.

I'd strongly encourage reading the full article; this is a case where reality has a surprising amount of detail and those details really matter. Many details are disease-specific, differing both between and within viral and bacterial infections, but were later assumed to generalize. Some are about measurement: researchers often used agar plates as a proxy for infected organisms without allowing sufficient time for airborne spread across a room (potentially hours) or with non-representative airflow, or imaging techniques which are inherently focussed on short distances and do not show the smallest particles. Others include conflation of particle transport through the air and particle deposition in the lungs, misattribution of the effect of spacing rules to distance (via droplets) rather than room occupancy, neglect of size change due to de- and re-hydration, and many more.

Ultimately, I think the question should not be "What's wrong with the virologists" but with the field of virology, and it seems basically normal to me: researchers thought they understood what was going on, disconfirming experiments were nontrivial to conduct and older research was generally ignored, and nobody influential wanted to listen to outsiders with awkward questions.

Contra the other answers, I contest that virologists haven't been running studies like that because they believed they had already been run.

And they are correct, studies had been run, from the 1940s to the 1960s. You are also correct - the setup you propose is exactly the one which cracked the problem for tuberculosis. Those results were published in 1962 by William Firth Wells.

The problem is that the WHO mostly deferred to the CDC, and the CDC guidance was accidentally written for weaponized airborne transmission rather than airborne transmission in general. So everyone, virologists included, believed the question of airborne transmission was largely settled and 5 microns was the magic number for airborne transmission in general, when in fact that was the magic number for bioweapons which can bypass our mucous lining instead. In addition to this error, the head of the CDC at the time the 5 micron number was recorded had for most of his career criticized the group of people who made the airborne transmission discovery, and to make matters worse Wells died in 1963 so there was no persistent authority to correct the matter. Also worth mentioning is that Wells was an engineer, and not a virologist.

The only people who had cause to question the "established facts" are people who a) had deep domain expertise that applied to the physical problem or b) were highly sensitive to the contradictions provided by the evidence.

I feel like this is basically expected, since the levels at which we expect most virology to work won't encounter this naturally. Wrong disease prevention guidelines won't show up if you are studying the shape of a virus, or its impact on tissues, or its genetic code; they also won't show up obviously in big picture characterizations like R number because these include multiple factors without making clear what they are up front.

In summary, virologists weren't doing the studies because they thought they had been done and we had our answer, but that answer was wrong in the exact same fashion as a wrong entry in a published steam table. In my view the problem is not with virologists the practitioners or virology the field, but with the fact that a bunch of bureaucrats were using a wrong reference. The reference would normally be corrected, but the best person to do so died too early and was a disparaged outsider to boot.

I sourced most of this from this old linkpost from 2021 and my own spot-checks.

I'm not a virologist, so this is a guess from the outside.

Most of modern science works on question substitution. You want to know whether A causes B, but this is hard to measure, so you look at whether A and B are correlated in observational studies. You want to know whether or not A is related to B, and so you look at the probability of observing what you did if they were not related, and conclude they are if that probability is sufficiently low. And so on.

As well, most of modern science is determined by a three-sided market, where both grant-makers need to like a proposal enough to fund it, scientists have to consider a study interesting/rewarding enough to run it, and journals need to consider a study interesting enough to publish it. Would a straightforward experiment as you describe gotten funded, run, and published?

In that background, my main hypothesis is: they already had an answer to a question that seemed related enough that they pretended the questions were the same and thus the answers were the same. [See this article; it wasn't until during the pandemic that someone traced down why virologists believed what they believed about airborne transmission, and also details the difficulties getting published of someone trying to fix broken beliefs about this sort of basic science.]

There's a simple, terrible answer:  because studies are hugely expensive, very time intensive, take a very long time to complete, and require multiple very slow iterations to get everything through committee in a way that our institutions will accept.  Consider:

 - Nobody is funding it.  The cost is literally hundreds of millions of dollars to do in a way that the medical establishment would accept.  Even then it would be challenged.

 - It would take thousands of man hours.  Ain't nobody got time for that.

 - It would take 3+ years to get everything approved and done properly, otherwise the medical establishment won't accept it.  Actually they still probably won't.

 - By the time you're done, it's a virtual certainty that the virus will have run its course and the result will be useless.

IMO, the above is more than sufficient.  The incentives were not there - or rather, the incentives were not sufficiently large to justify the cost and were further derated by the expected utility of the information a year after the pandemic is over.