The problem is trying to compare apples and oranges. When people are lead to believe or think they know that Narrow bands will not measure this that or the other thing, when the truth is they have little to No idea what they can do. My whole point has always been the same and that is simply that the Broad Band systems on the market today are not capable of even coming close to what is claimed. The Broad Band sensor is capable of being much better BUT due to how they are being used they are not. So if the poor range they have today, due to the misuse of them, is good enough, that's fine.
Now if you take a simple Narrow Band sensor and use it, it cannot measure the entire range of a Broad Band can do. If we take the range that we need for most internal combustion engines running gasoline we DONOT need the entire range of a Broad Band sensor. A Narrow Band can and will do the job well enough! Not as well as a properly used Broad Band can, but, well enough for the applications we are talking about and as well as the way Broad Bands are being used today.
Another little missed fact is that the response time of a Broad Band sensor is ~280 Ms once up to operating temperature, that's as fast as it can respond and there is not a damn thing you or I can do about it! So that gets you 3.57 samples per second MAX. As they age or measure richer mixtures that slows some more. An engine fires each cylinder 8.333 times per second at 1000 RPM, so a Broad Band gets you (8.333/3.57) 2.33 engine firings per sample taken at 1000 RPM. Now let's look at 6000 RPM! That's 50 engine firing cycles per second, so (50/3.57) 14.00 engine firings per sample taken. So a little simple math shows you just how much your not seeing!
A heated Narrow Band sensor up to temperature is ~10 Ms today! So that gets you 100 samples per second MAX. In the HD case the sensor is read once each and every firing cycle at the same place in crankshaft position each time. So, An engine fires each cylinder 8.333 times per second at 1000 RPM, so a Narrow Band can get you (8.333/100) 0.083 engine firings per sample taken at 1000 RPM. Now let's look at 6000 RPM! That's 50 engine firing cycles per second, so (50/100) 0.5 engine firings per sample taken. Since HD has limited it to once per engine cycle it's easy to see the sensor response is not at issue at all with the Narrow Bands like it is with Broad Bands.
So with this information which would you want to be using?