Dear All:

Thanks for all the interesting and thoughtful comments in response to my provocative posting. I'll address them in a moment, but first a few clarifying statements:
(a) Of course, like you, I also enjoy driving in "stealth" (pure-EV) mode when it occurs. And, it really does impress non-Toyota-hybrid owners. The question is whether trying to force pure-EV operation as much as possible actually harms, rather than increases, fuel economy. That's what my post was about.
(b) Yes, recovering kinetic energy that would otherwise be wasted as heat in friction braking is desirable, and helps to improve the TCH's fuel economy, especially in-town. This regenerative energy recovery comes for free because one has the MGs present, but it would not be possible without having the NiMH battery in which to store this recovered energy. One wouldn't want the NiMH battery's state-of-charge to get so high that there's no more room left for regenerative braking energy storage, and indeed the TCH is careful to always leave some headroom for this. But, don't forget that this recovered kinetic energy originally came either directly from the gasoline via the ICE, or else was imparted to the car by the NiMH battery and the MGs. However, the electricity in the battery (whether regeneratively recovered or not) ultimately also came from the gasoline engine. So, ALL the car's energy sources ultimately point back to the ICE.
(c) The principal use of the NiMH battery's stored energy (apart from starting the ICE and powering the air conditioning) is to add extra on-demand torque to the ICE's limited output when the driver (or the road) requires this. This is only possible for brief periods because of the battery's linited storage capacity. But it does enable the ICE to be designed for maximum efficiency (Atkinson-cycle) without the associated reduced power output capability being a liability under normal circumstances. [This small 4-cylinder Atkinson-cycle engine's power would be inadequate in a regular non-hybrid Camry.]
(d) Toyota's control system attempts to maximize the efficiency of the whole system by trying to always run the ICE in the range where its efficiency is greatest, and letting the battery supplement the ICE's output when necessary. During acceleration the battery helps to supply power to the wheels. This is when the ICE is forced to run at less than its highest efficiency. Once steady cruising speed is reached, and assuming a level road, the ICE will be running much more efficiently, and this is when the NiMH battery is recharged by the ICE. So, the electrical energy sent to the battery by the ICE is in this sense already "higher-efficiency" energy. Moreover, the ICE is FAR more efficient than the battery as a source of motive power at cruising speed, because of the losses (guesstimated at say 34% in my previous post) associated with the multiple energy reconversions from ICE to battery to wheels.
(e) To do better than the built-in control system, one would have to generate electrical power more efficiently than Toyota has been able to program the system to do. One would have to use battery power ONLY when it would be more efficient than the ICE (taking into account the 34% loss associated therewith), and use the ICE ONLY when it would be more efficient than the battery as a source of power (again taking into account the associated 34% loss). Since this must be what Toyota is already attempting to do, I say that you have a very difficult (maybe impossible?) task ahead of you!
(f) It follows that the greatest possible fuel efficiency achievable by the TCH is that of its ICE ALONE, running at that speed at which it is most efficient at producing the needed power. You'll never exceed this on average no matter how hard you try!

I am now ready to attempt to respond to your comments:

wcmack — AsI have argued above, you CANNOT in the long run exceed the efficiency of the Atkinson-cycle ICE on its own, since it is the source of all the power, and so I think that the HSD cannot overall do any better than the Atkinson-cycle ICE. In your first 50-mile scenario, I agree with you that both vehicles would achieve the SAME FE. In your second 50-mile closed-loop scenario, I agree with you that the TCH WOULD do better. If the NiMH battery's state-of-charge is the same at the end of the circuit as it was at the start (i.e., if the "ups" use as much battery energy for MG torque boosting, as is recovered regeneratively in the "downs," and in principle this WOULD be the case), I think your argument is sound.

lars-ss — As long as the vehicle runs in pure-EV mode (taking into account the associated 34% loss) ONLY when it is more efficient than using its ICE, then it WILL benefit from doing so. This is presumably what Toyota is trying to achieve with its system control, and why the TCH DOES go into pure-EV mode when it decides that it wants to. The battery is mainly intended to store recovered kinetic energy and then use it for starting and torque boosting the ICE when necessary. I believe that most other uses actually cause overall FE losses. Any energy placed in the battery has incurred a net 34% (say) penalty by the time it is used. I have argued above that its most efficient use is for torque boosting, not for powering the car in "stealth" mode.

greenvillehybrid — The traction battery is FREQUENTLY used for torque boosting the ICE, especially when accelerating. The energy-flow display clearly shows this. That's where your recovered kinetic energy is best used. You're NOT wasting the energy of regenerative braking by not running the car in pure-EV mode. You have a better use for it. After all, the kinetic energy was produced by acceleration in the first place (that's what got the car up to speed from rest!), and so it's fitting that it be used to assist in getting the car back up to speed again. You're right that the traction battery must keep "headroom" available for storing regenerative braking energy, and indeed it does do this. The fact that it stays approximately fully charged all the time probably reflects a more relaxed driving style and/or your local terrain.

Joe.B — You're right of course, and I sincerely hope that my original post did not leave the impression that there's no point in storing and recovering energy in the NiMH battery, even though there are losses associated with this process. This energy recovery and re-use is a MAJOR advantage of a hybrid design. This is particularly valuable in in-town start-stop driving (and here the fact that the ICE shuts off frequently is also a MAJOR benefit — and don't forget that it's the NiMH battery that is used to restart the ICE each time). But since all this recovered energy came originally from the gasoline via the ICE, the ICE's efficiency alone sets the limit to what's achievable overall. Yes, indeed, use this electrical energy; but running in pure-EV mode is usually not the best use for it.

turk — (1) Yes, I agree with you — you certainly DO want to use the energy recovered from regenerative braking, even though there are losses associated with this recovery and re-use. Without a MG and battery this couldn't be done. I was not arguing against this. (2) Please see my reply to wcmack above. (3) Agreed. All in all, we're not in disagreement, but I hope that my introductory comments above will help clarify things for you.

Pete4 — Regarding your point that it's best to deplete the NiMH battery before shutting off the TCH at the end of the day, so that the ICE will do useful work by recharging the battery during its initial warm-up the following day, this seems plausible; and if this is when you try to force pure-EV mode, I could be convinced by your argument. Many TCH owners, however, appear to be trying to force pure-EV mode whenever possible. And, yes, the ICE is more efficient under heavier load, and that's the way Toyota use it! It's what I was getting at when I said that they attempt to run it so as to produce the needed power in the most efficient manner, and the HSD allows them to do so. I have been using a ScanGaugeII (www.scangauge.com) to monitor the ICE load factor for the past month, and it's usually high whenever it's running. Battery charging done when it's at its most efficient is best. This WAS all incorporated in my original post, and my elaborations above may make this clearer. Sorry if I wasn't as clear as I would have liked to have been! As regards "pulse and glide," I'm skeptical that this really does result in a FE improvement overall. Can you document this convincingly? I think that my argument shows that this cannot be the case, because of the losses associated with the multiple energy conversions involved.

Well, that's all for now. This is becoming an interesting thread. Sorry about the length of this reply!

Stan