I don't know that I have any specific advice that hasn't already been covered, but looking at all the different stuff going on here, I thought I'd share a few of the weird things I've seen.
- I had one site where I was to replace a failed IP dome camera attached to the Alucobond sheathing of a building (Alucobond is a sheet building material comprised of a light-weight plastic core with thin aluminum skin on either side, making it conductive). I removed the old dome (an IQ-A11), connected network and power (12VDC) to the new one (rebranded HIKvision) and went inside the store to ensure it was working and configure its IP. At that point it was working fine, so I went back outside, mounted it to the wall, went back inside to check the view... and found signal loss.
Thinking maybe a wire had been pinched, I went back out, pulled the camera off the wall, checked all the wires, then went back inside to check its function - working perfectly. Went out, re-mounted it, headed back inside... signal loss again. ARGH.
I went through a few other steps I don't recall, until I finally discovered the camera shut down as soon as it touched its metal mounting plate that was screwed to the wall. On a hunch, I pulled out my multimeter... and got a reading of some 40VAC between the camera body and the mounting plate. Something was putting a significant potential into the entire outer wall of the site, relative to the camera's power supply (which if memory serves was a transformer-and-regulator type, which SHOULD keep the power fully isolated).
In the end, I made it work by cutting out a piece of cardboard to put between the mounting plate and the wall, and rubber O-rings to isolate the screws from the plate, and it's still working today, 3-4 years later. Moral of the story: a potentially damaging voltage differential can exist almost anywhere... and more to the point here, can indicate a far more serious problem somewhere else. And yes, I reported the issue to the site and noted it on the service order for the building maintenance outfit... whether anything was ever done about it, I don't know.
- I've often had ground-loop issues using video baluns and Cat-5e, and on one site, it finally got to me badly enough that I took the time to figure it out: the problem exists when you put cameras with shared power and video grounds on a power supply with a common power ground for all those cameras.
The balun essentially puts a coil in-line with the video signal and video ground (two coils, actually, one at each end), which adds a few hundred feet of length to that run, as well as several dozen ohms of DC resistance. When multiple cameras have a common power ground, and a common video ground (at the DVR) that's several times the length and with a lot more DC resistance... well that's a textbook recipe for a ground loop.
In my experience, the majority of 12VDC-only cameras DO have a common power and video ground internally, whereas dual-voltage AC/DC, and 24VAC-only cameras use an internal rectifier and regulator that isolates the power ground from the video ground.
That might be somewhere to start for your ground loop problems... not that it's a priority, BUT...
As others have suggested, there may be a causal correlation between the ground loops and the camera failures - ie. the two issues have a related source. One possibility (just thinking out loud) is that a relatively low voltage, say from a leaky power feed to a remote building, is being carried in one of the conduits, and then being carried into a camera mounted to a metal pole... that may then be felt by other cameras via shared video and/or power grounds (or a combination of both!). That potential may not normally be enough to cause an issue, but suppose a power "brownout" (common in rural areas) means an INCREASE in that potential relative to ground, which means a surge fed into the grounds of all those cameras.
Again, this is purely slightly-knowledgeable speculation... but now suppose the original cameras all had solid enough power supplies and video systems to withstand these fluctuations, but those replaced after the lightning killed them do not? That might explain why six of the eight new cameras have subsequently failed where the old ones keep working (note: it might help to know what brands and models both the old and new cameras are).
As for the source of that stray voltage, it's possible the lightning strike damaged the insulation on some of the camera wiring OR SOME REGULAR 120VAC WIRING that is now creating a potential in the metal and then in the camera wiring... that's something for the electrician to sort out, but it's easy enough to test by pulling a camera off a pole, then measure for voltage (AC and DC) between the camera housing and the pole.
Personally, I'd start using all that coax to fish through new Cat-5e so as many cameras as possible can have home-run power instead of power supplies scattered all over the place (with the added bonus of easy IP upgrades later). As has been suggested elsewhere, I'd also look at isolating the cameras from their earth grounds (aka metal poles).
All this, of course, is in addition to the electrician making sure there are no problems with the power source, and adding UPS units where possible to clean up the power going to everything (another benefit of home-running as much power as possible is to get more cameras on the filtered source).
Plus if you want to add lightning arrestors, which by nature need their own earth grounds... maybe THOSE could simply ground to the poles, which by their nature are well grounded. (Note: I've only ever worked once on a system that needed serious lightning protection, so I can't say if that would be sufficient grounding... someone with more experience in that area can correct me if necessary.)
I know it's a lot to read, but hopefully it helps :)