Field strength is field strength, no matter the modulated occupied bandwidth or frequency. The difference in receiving a signal at any distance is because of atmospheric and reflective propagation for a particular frequency range. It has nothing to do with deviation or occupied bandwidth, let alone at audio frequencies.
I realize this is yet again the same old tired (and wrong) argument you keep bringing up Bruce, insisting that stations broadcasting HD-FM are somehow degrading their field strength by occupying more bandwidth. Not only is that premise completely incorrect, but as several of us here have tried to explain, most HD carriers are transmitted through a second transmitter and in some cases, a separate antenna than the analog station. For those stations that are using a single final amplifier, the amplifier is broad-banded to maintain the correct licensed TPO for analog and digital carriers.
Again, how do you explain a DTV station that occupies a 6Mhz with constant digital modulation? It's going through a broadband RF amplifier?
Let me address the DTV question first. In the analog days, there was a whole line of op amps that were optimized for flat differential gain and flat differential phase out to 6 MHz. The Base part number was THS4222. It is still active, but most of the references to video applications have been removed. I also notice that although the dual part is still made, the triple channel part - the THS4227 - is not. You would ideally want a triple channel part if you wanted a separate amplifier for R, G, and B. If you look in the spec table for the THS4222, it still mentions differential gain and phase for NTSC and PAL - although the part is now marketed as a telecomm grade A/D converter with excellent HD2 and HD3 response. No - that has nothing to do with HD radio! It is second and third harmonic distortion - but I couldn't resist putting the spec in here because the nomenclature is identical to HD radio!
DTV is probably fairly immune to differential gain and phase distortion - although the THS 4222 would still be an excellent choice to drive digital TV signals. The run of the mill op amp is optimized for 600 Ohm, but the THS4222 is optimized for 150 Ohm loads (75 Ohm load and 75 Ohm series matching resistor, because the output impedance of an op amp is low). I notice they also show the THS4222 as a differential line driver - I believe it would be excellent for 600 Ohm drive in studios, because of the hefty output stage.
Enough of the video stuff, interesting, but not radio related.
I agree that field strength is field strength, but I made the mistake when I was doing drive tests 35 years ago of using a spectrum analyzer. It had several bandwidths, but fixed gain. The fixed gain part is where I made my mistake. A consumer receiver is non-linear - it has AGC. So there is no difference to the consumer if the radio is getting 10 dBf or 100,000 dBf. Except maybe static on the low end. There are several things going on with HD radio. I am tending towards YOUR viewpoint on the antenna, and now think Houston's FM coverage problems are related to a poor choice of antenna bay and not necessarily the sidebands. And - thank you for being so hard nosed about it. The dangers of jumping to conclusions! HD on - 60 miles less range than HD off. Must be HD, right? No - not if they also changed the type of bay and the bays they are using for HD suck. That would explain why KBPA is such a blowtorch over Houston from 150 miles away, and they are running HD. They are doing something really different, and RIGHT - as opposed to something Houston is doing WRONG. Someone told me Houston stations switched the brand of bay when they started broadcasting HD. And if KGLK had the old bays active when they weren't using HD, and the new bays active when they were - there is a plausible explanation for the 60 mile difference! I don't think they care about Centerville, Madisonville, and Huntsville - but they darn sure care about building penetration, and if they lose 60 miles of range, something dramatic is probably happening over Houston where it DOES matter. I'm not the engineer. But if they seen this post, attention needs to be paid to the signal! You and I might disagree about whether or not it is HD or bays, but the range difference was observable, and repeatable. That is empirical observation - which needs an explanation as to the cause.