With a full charge the gauge has always dropped 1-2 bars after several hundred miles and EV range will drop a few miles. This may be part of the design, to allow fully charged batteries to discharge a bit to accommodate regenerative energy.
You didn't state it in your initial example, but I was assuming that HV held at the initial set-point in both cases.
I did state that a fully charged battery will show a drop in SOC/EV range after driving for a period of time in HV.
Theoretically, HV maintains a specific SOC once set. In practice it allows some discharge from a full SOC. I have observed an initial drop of 1-2 bars within the first 100-200 miles. After that drop it has held quite steady, even after an additional 500-600 miles.
Some data has indicated that the battery will discharge ~4% before recharging begins in HV. My observations support that, in that EV range will fluctuate depending on power demand. Some owners have reported a steady decline in SOC when operating in HV for extended periods.
I believe there are two main reasons that explain the gradual reduction of SOC while HV is engaged.
1) There is a 99% chance that the vehicle will be turned off with the SOC below the set point. The SOC is constantly fluctuating between the set point and 4% below the set point. Unless the ICE is allowed to run until the set point is reached, some discharge will occur at every stop.
2) The default operating mode is EV. If HV is not engaged immediately upon starting, additional discharge will occur.
Ultimately, in our PHEV, we can easily determine HV range by operating exclusively in HV. SOC may taint the results by up to 2%, which I would consider within the margin of error. We can also easily determine actual miles travelled in EV, provided the ICE does not engage due to excessive power demand or a system check.
One thing we can’t easily determine is mpg with the ICE, because we cannot operate exclusively using the ICE.
