Did you ever figure this out? I am on my second run with my herms setup and I'm getting terrible efficiency... I do think part of it is due to my mash profile I'm using with beer smith however I think I still have quite a bit of fine tuning to do with it. My house ipa was a whole point off OG and my Belgian pale ale I just brewed was .6 points off...
When I first brewed on the new herms I was short. Maybe this will help.
In a nutshell, my MT draws from the bottom using a strainer tube. Pockets of grain, several pounds of it, sit in piles on either side of that pick-up screen. I either need to build/buy a false bottom, or make up a concentric smaller grain cylinder that rests inside of the main keg. That way 100% of the wort flow would be through the grain cylinder with no channeling, no direct route to the pick-up point. Watching how the Braumiester 550l system works (youtube), got me to thinking how inefficient my process previously was, so I started stirring those areas at the bottom of the keg and to the sides of the screen tube several times during mashing. That way those grains were occasionally reincorporated into the main mash, the conversion process and ultimately more sugars getting dissolved into the wort. Wallah... all of the sudden I was beating recipe brewhouse efficiencies by 4-5%.
So soon, it's time to fab up a false bottom that allows no bypass and holds the grains high (a few inches) and away from the pick up, but it's too hot now for metal fab, it's boating season, there are bikinis that need patrolling and the keezer is full, so maybe in the fall.
No idea how or where your PID measures temp (mine is at the herms coil discharge point, where the warmed returning wort re-enters the top of the mash tun). This may be a factor as it was for me. A HERMs keg system may have 25 total gallons of wort/water during the mashing process. It takes time for the entirety of the water to reach temperature. Just because the HLT and the wort passing through the coil is at selected temp, doesn't mean that the entire MT's contents are. We're measuring little areas with our thermocouples. You need to factor that data into a macro system view, as any one point of measure is just a limited glimpse of the whole system temperature. Kinda like pushing the car's accelerator does't immediately result in the desired speed, unless you're driving the new Tesla X.
As an example, at mashout I was reading 168, I'd start a 10 min timer and start to fill the boil kettle, but that measure was only the small trickle of thick wort returning to the cooler top of the MT. The failure, much of the MT never reached 168, maybe it was 10 or even 12 degrees cooler. Consequently, my first few batches never really got any of the benefit of that final rinsing of grains with hotter water. Pre-setting the mashout temp about 10 minutes now, prior to starting the time that I wanted to have hotter 2nd running and sparge water, also helped with efficiency gains.
As a general rule mashing for 75 mins. vs. 60 in most cases has also resulted in gains, mostly compensating for this "system accretion lag".