Xu's other recent works has been lacking meat and are better at fear-mongering (Tesla ultrasonic 'hack', Smart grid 'hack').
The course of reasoning I see proposed is:
-> Personnel in Cuba received an odd class of injury
-> High frequency sound can cause this class of injury and seems the most likely candidate
-> HF sound doesn't propagate well in air and has many other properties that make it a poor tool for espionage/malicious actions
-> Xu's work demonstrates that sound interference from two HF sources can create unexpected effects due to non-linearities in the physics behind sound/air
Xu's work is the last piece in the puzzle IMO. It offers a plausible scenario: the building had Cuban tech and when we moved in we installed non-Cuban tech that, due to some benign like motor speeds, creates sonic problems for building occupants. The reason this doesn't really happen anywhere else is because of the unique situation surrounding the building.
The next steps I'd take were I investigating this is to look at what equipment was installed at the facilities and then determine what frequencies the equipment operates to try to find a likely candidate.
Edit: I realized you may not be familiar with equipment noise problems, these two articles described problems in the infrasound range.
[0] https://gizmodo.com/some-ghosts-may-be-sound-waves-just-belo...
[1] https://www.theguardian.com/science/2003/oct/16/science.faro...