OPUS Weirdness with Ortho Hgts

Have a project to monitor water levels on a half dozen lakes and a contractor established NAVD88 elevations on benchmarks at each location. I'm going out a verifying the elevations and VBMs that were set before getting water level readings and I ran into something that has me a bit stumped. Hoping someone might have some idea of what's going on. 

The outfit that did the work established an elevation of 80.86' on an Al cap on site using OPUS. I happened to have a nearby control point (.5 miles) from some work I did a couple of months ago that I had an OPUS on and great sky and I used it to RTK in and check the elevation holding the elevation on my control point. I get an elevation of 80.16, so about .7' difference. Okay. So I do another fresh set up using my control point and do 10 minute PPK obs on the same point, get a fresh OPUS on my base to use, and get almost exactly the same result for both my base point and the AL cap in question. At this point I'm more than a little stumped, so I do a direct static on the Al cap in question for 2+ hours, and get an ortho height of 80.72'. The cap in question was set a couple of years ago by another surveyor I know and they forwarded me their control sheet for it which shows an elevation for this as 79.94'.

As a sidenote, I've matched up near perfect on a couple of the other benchmarks of theirs I've checked.

Next week I'm going to borrow a couple of other receivers and do a simultaneous obs on both for 2 plus hours. I've checked everything I can think of and still don't know why I'm getting this .7' discrepancy. Any ideas? Trimble gear, R780 and R12i, using Access and TBC for the post processing. The only thing I can think of is it might have something to do with the CORS stations OPUS is using? Still, seems like a lot. 

TIA and have a great day.