The real question is whether it's a display quirk--we all know that the guess-o-meter (GOM) is not very accurate. I think that many of the complaints of "HV charge doesn't actually hold" is people noticing their guess-to-mated range going down while using HV (when HV is likely [trying to ] keep the system-of-charge (SOC) the same [and sometimes struggling], but letting the GOM reflect the current driving situation for that SOC).
My conjecture (probably wrong):
- The engineers let the GOM go down in HV mode when the driving conditions get tougher (e.g., when climbing) because to allowing the GOM to overestimate range would lead to people running out of range, but
- The engineers may have coded the system to have the GOM underestimate range when driving gets easier, so the GOM does not go significantly above the range that existed when you hit the HV button, because allowing it to exceed this range would confuse people. If you're actually banking power (e.g., going downhill) the actual SOC may be going up which is not being reflected on the GOM
Seems we could test this conjecture with an appropriate ODBC reader...(or perhaps with a restart after the cycle and reread the GOM)