I think it's a little more complicated than that. Current battery technology is pretty impressive with respect to efficiency (something like 99%). Generators and alternators are pretty efficient - they need to be since the vast majority of electricity is produced by converting some form of rotational kenetic energy into electricity (maybe in the 85-90% range?). Internal combustion engines have notoriously poor efficiency (so there's significant room for improvement) and their efficiency is a function of RPM. Although the continuously variable transmission can adjust drive ratio to get the RPM for the best efficiency of the ICE at the current HP requirement, the RPM may still be at a point where higher HP (and corresponding RPM) would be more efficient.
I haven't been able to find actual efficiency numbers on any of the components in question. Even if we had them, HP requirements are going to be a function of speed, road, and even rolling resistance of the car. I wish the data was published, but apparently it's a secret.
With those unknowns, the car might be cruising along at a HP requirement where the resulting RPM corresponds to a point on the efficiency curve where higher RPM would result in lower efficiency or higher efficiency. Suppose it's a case where the efficiency of the ICE when operating in normal HV mode works out to 20% but with the higher load from HV charge mode the CVT can push the RPM up to a point where the ICE efficiency was 22%. If the batteries are 99% efficient and the generator/alternator is 80% efficient, you're ahead.
The problem can get even more messy if you expand the system. Suppose a tuck-under garage and an air conditioned house. Traveling the last 5 or 10 miles on electric would allow the ICE to cool before parking it in the garage, which reduces the heat dissipated into the garage and then into the house - which is then pumped out of the house via the AC.
Basically, without real numbers for efficiency of the interesting components, all someone can do is gather some empirical data and see how it looks.
I agree that in situations where running HV Charge puts the engine higher on the power curve than it would be otherwise, that this could more than offset the inherent losses that occur when storing and retrieving electricity in the battery. There are also mechanical losses from using ICE to turn the starter/generator motor to create electricity, as well as losses when the electricity is used to turn the traction motor, however that does not really factor into this equation because HV mode does exactly the same thing.
Except when running in "gear mode". That's where ICE is connected directly to the traction motor via a clutch, and since the traction motor is essentially directly connected to the wheels, then in effect ICE is directly powering the wheels. It uses a fixed overdrive gear in this situation, Clarity does not have a CVT. When not in gear mode then ICE only turns the starter/generator motor to create electricity to be used by the traction motor. Even as efficient as those two electric motors may be, it cannot be as efficient as when ICE is directly connected to the wheels with an overdrive gear. I bring this up because I am not sure if HV Charge ever uses gear mode, in which case regular HV mode would have a slight advantage over HV Charge, at least in this one aspect. Then again maybe HV Charge does use gear mode at times, I will have to pay attention next time I use HV Charge as I don't remember. By the way for anyone who doesn't know, you can tell when ICE is running in gear mode by the tiny gear icon that sometimes appears on the energy distribution display. Gear mode is used only when driving at steady speeds above 45 MPH or so.
Some people who have done tests have reported that HV Charge can at times be more efficient than HV. If so then it could be the power curve situation that you are describing. But it's so hard to measure any of this. The only way to accurately measure MPG is to calculate it based on gas purchases and miles driven. But that is only meaningful if the entire tankful is driving in HV. Even then you never know if some of the electricity used in HV mode came from a previous charging session. So you really have to start the test with a depleted battery.
Anyone using the MPG displayed on the instrument panel to figure any of this out is wasting their time as that is notoriously unreliable. Gas receipts and odometer readings are the only meaningful measurements when it comes to MPG.
Not that I plan to do any of these experiments. I just accept other people's reports that HV Charge is similar efficiency to regular HV. I mean people who have done actual testing, not people who just look at displayed MPG on the instrument panel while running in HV Charge, and don't even factor in the value of the EV miles that are generated.
If there is a situation that will make the driving experience more enjoyable by using HV Charge to build up some EV miles for later use (city driving, climbing a hill, etc.) then I have no hesitation to use HV Charge. One person reported that they use HV Charge instead of HV because it tends to result in lower overall engine noise. I realize that is counterintuitive since we automatically assume that HV Charge will add RPM on top of whatever RPM the engine is already doing, but that is not really the case. It seems more like HV Charge does "opportunity charging" during periods when ICE is not being used (slow speeds or stopped) or if ICE is being used at low power. The resulting additional battery charge is then available to help avoid the high RPM's that can occur when the battery is depleted.