The 'answer' is she did NOT spend anything to 'build' those observatories, she and her team spent grant money to 'rent' them to make the observations, (
Not true: the uber expensive ALMA observatory data was provided gratis. They won free time on it on the merits of their proposal.
You can go through the "Acknowledgments" section of the Nature paper and look at the foundation and governmental grant money.
It's very small and compared to the tens of millions required for a spacecraft, practically free.
No one is criticizing the paper in Nature. In fact, I'm praising it because of the economy of their effort, and the
innovative modeling and methodology which is every bit as important as their putative discovery.
I believe Nature Astronomy has an acceptance rate of ~ 10% which is highly selective. Per their metadata,
it was a 5 month review process which is typical for highly selective venues.
But I'd wager that 90% of the papers that didn't get in were also very high quality; acceptance is decidedly
non-deterministic, non-repeatable, your reviewers are your rivals and it's a horrible thing on which to base a career (end rant).
PI salaries are typically paid by the university. Undergrads are regarded as "free" by research programs since you aren't
charged dept. overheads for using them. Grad students are more expensive but still cheap and you have them captive since
you are the gatekeeper on them graduating and staying funded*/enrolled.
Post-docs are a different animal but they want to build their pubs and get their letters of recommendations/job talk
connections. And their productivity per $ spent is vastly better that undergrads or grad students 'cuz they 1337;
getting a post-doc at a top tier group is harder that getting in as a grad student.
* unless you win external fellowships which have like single digit acceptance rates for the nicer ones (e.g. NDSEG).