OK, I've now seen the video. I'd not seen it before, so thanks for the link. The first thing to say is that Jerry Lang clearly does generally know what he is talking about - I don't doubt that he has a lot of knowledge about combustion.
He claims that burning of gasoline in air is relatively slow, and this leads to a big part of the burn occuring (uselessly) after the exhaust valve has opened. Supposedly the MPG-Cap produces free hydrogen in the mixture, which speeds up the burn and so prevents this waste. In principle this could be plausible, because hydrogen really does speed up the burn, but you would need a
lot of it to have any significant effect - and it is not at all clear to me how the MPG-Cap can generate this, or how FFi have proven that it happens. The analogy with the 100-foot tube of mixture is fairly misleading, because the burning speed in an engine is very much faster due to the high levels of turbulence.
Critically, car makers already know that burning speed is important for combustion efficiency, and so optimise the level of turbulence (and hence flame speed) appropriately. I, personally, have done many experiments on engines where I have speeded the burn up by 50% or so and seen
no measurable improvement in economy. While a faster burn might reduce the (in fact very small) loss due to late burning, it increases losses due to heat transfer to the cylinder head, plus a few other effects, leading to a very small or even zero overall benefit.
(Edit: it's also worth noting that there can be some very significant
disadvantages with a burn that is "too fast" - quite apart from the NOx emissions mentioned later, a faster burn would cause higher in-cylinder pressure, which at full load might lead to mechanical damage such as a leaking cylinder head gasket. On at least one project I have worked on, the engine output had to be artificially limited to reduce the in-cylinder pressure - speeding up the burn on this particular engine would almost certainly cause it to break. By the way, I have yet to see a single piece of evidence from FFI, or even a
reference to testing, where they have actually measured the burn rate in an engine and showed that it is faster when the MPG-Cap is used - although this is an easy and cheap test to do.)
Talking about race engines with flames coming out of them is a bit irrelevant, since race engines are not that similar to normal road car engines, and also operate at very different conditions (typically very high speed, which naturally reduces the time available for combustion).
Another key point is that, as I mention on my "turbulence" page (
http://www.fuelsaving.info/turbulence.htm ), the optimum ignition timing (or injection timing on a diesel) is very strongly linked to the burn speed. Basically, a faster burn needs later ignition to keep the bulk of the burn at the optimum part of the cycle - otherwise it will occur too early, when the piston is still on its way up. Even if you really could get an economy benefit from a faster burn, without ignition timing changes the result is more likely to be
worse efficiency.
The claim about NOx reduction by faster burning might be correct - I'n not an expert on NOx emissions - but does not fit my experience with variable-burn-rate engines. Claims of 75%+ reduction also do not fit the around 20% improvement seen on the Australian test discussed earlier in this thread.
Then Jerry talks about his experiments to test economy benefit "in the real world". This is a perfect example of how
not to do testing, since it is not blinded and there is no placebo control. Every one of the test subjects was doing something different to normal, and knew they were testing "something", so it's not suprising there was an effect. It would have been trivially easy to randomly assign the drivers to get either the pill or a placebo and then compare the difference between the two groups, but FFI - yet again - failed to take this basic step.
Interestingly, Jerry then mentions the Millbrook tests and
admits that the cap failed the test. He makes some claim about Millbrook using an inappropriate "modifier" to the data, which makes no sense to me based on my knowledge of this type of testing. He also claims the test used less gasoline, but did not show any benefit on the calculated economy number. Since the calculated value is the
only measure of how much gasoline was used, this sounds like B.S. to me. Crucially, although Jerry has a lot of experience of industrial burners, I believe he is new to the field of car economy and emissions testing.
I would love to discuss some of these points directly with Jerry, but doubt that will ever happen.