Ubiquiti hardware and high power HF

What’s the issue?  And what is “high power” ? 600? KW? 1.5?

This hardware is being installed at commercial broadcasting sites by WISPs everywhere. The only issue I remember reading on the Ubiquiti forum a while back was RF from high power FM stations near 100MHz getting into the Ethernet port when running at 100Mbps. But that is not the case here.
Sounds like your friend has an issue with RF in the shack, and most likely it is affecting other devices too, and possibly his neighbor's. I would suggest start with replacing that G5RV with resonant dipoles for 40 and 80m.

Are the Nanobridges shielded?
Is the Cat5 shielded?

I'm going to combine a few different replies into one here:

As noted these devices have been installed on AM/FM radio towers, but this could be a frequency dependent failure if we are hitting an oscillator inside.

First off there is a note for using shielded cable, I would actually up this to make sure your using double shielded cable since we are talking so high power here. Double Shielded has both an aluminum shield and a braided high current shield.  However there is also an exception to this that it might also be better to run fiberoptics up to the node and convert at the hardware to Shielded Twisted Pair (STP) wire if the cause is because of induction on the  network cable.  This has been done on some FM tower installs because of issues.

Ferrites: Depending where you put them I would wonder if they are not actually adding to the the problem,  the Ubiqutii hardware relies on the shield as part of the drain path to pull EMI away from the device, if you choke the cables at the ground side you may be forcing it back into the hardware.  On top of that if you are a lightning arrestor for the network cable make sure it ties the shield to ground either at the arrestor or down stream at a good ground point on the shielded cable (many arrestors I've used especially from L-COM actually float the ground and rely on the output port to ground it unless there is a lightning strike where it arcs over the gas tube, while this is good if you have a solid ground at the computer or power injector end of the cable, if not it basically negates the shield drain purpose)

https://www.rfarmor.com/unbk5.html (previous link was a nanostation)  of course this will do little good if the direction you point the node is directly into the antenna but if the antenna is to the side or even better behind the node this can add some significant isolating keeping the RF power from getting in via the plastic case.

If that doesn't work I know some support for the Rocket Titanium is working its way through code review , it may be the better option (the Rocket Titanium is enclosed in a metal enclosure) once it hits.   1500w at 20feet is roughly 500v P-P (assuming unity gain on both sides) being thrown at the Ubiquity hardware. I don't know many ham radio operators who would put their TV that close to the antenna,  the same is true for AREDN gear without shielding, its a complicated device.

I'll admit there isn't much data on HAM frequencies being used near these devices, the oscillators would be my main concern of being overloaded and locking the device up for you.