Hi Carl.
I am curious where you read that two antennas are not a good idea. I'm NOT suggesting that they ARE a good idea; just curious. I suspect the answer is "it depends."
I'm a ham (Extra, AF7BC), and I know a bit about this stuff, so I'll throw out a few generalities...
The OUTGOING range depends upon how much of your signal ends up at the other end of the path (to the other guy). The INCOMING range depends upon how much of the other guy's signal ends up at your end of the path. Over relatively short distances (a few miles) the path losses are essentially the same in each direction so the differences (if any) will lie with the equipment (e.g., radios, antennas, etc.) and conditions (e.g., power lines, other, nearby, radios, etc.) at each end.
The amount of YOUR signal that ends up at the other guy depends upon 1) the amount of power you are putting into your antenna system, 2) the losses in the antenna system, 3) the way your antenna system distributes (spreads or focusses) your signal in space, 4) antenna height (which is sort of related to 3), above) and 5) environmental factors in the path such as ground conductivity (e.g. earth vs. water), obstructions, mountains & valleys (signal reflection), trees (signal absorption), atmospheric conditions (for a LONG path), etc. For a short path, up to, say, a few miles, atmospheric conditions can pretty much be ignored.
The amount of power you are putting into your antenna system is pretty much fixed: a legal CB AM (Amplitude Modulation) radio will deliver a MAXIMUM of four (4) watts (of carrier power) to the antenna jack of the radio (see Rule #10 at the link
http://www.cbradiomemories.com/fcc_rules_cb.htm#Rule%2010). If you have an SSB (Single-SideBand) radio the limit is 12 watts PEP (Peak-Envelope-Power). If you have an SSB radio, then you each have to have one. SSB radios have a farther "reach" but they cost more, and are usually "base station" type units. For example, I have never seen an SSB walkie-talkie.
Not to go too far down the tech-nerd path, but the sideband power for a four-watt-carrier, 100% modulated, AM radio is one watt for each set of sidebands, which is on the order of one-tenth of the sideband power of a 12 watt PEP SSB radio. And, the sidebands are where the information (voice) lives.
The losses in the antenna system depend upon 1) the LENGTH, KIND (RG-58, RG-8, etc.) and CONDITION (pinched, cut, abraded, kinked, etc.) of the coaxial cable between the radio and the antenna and 2) the losses in the antenna itself.
To illustrate cable losses, about half of your power would be lost to heat in a 100’ length of RG-58, before it even got to the antenna (
http://www.w4rp.com/ref/coax.html). This is just to give you an idea of the effect; you probably wouldn’t be using 100’ in your installation, maybe a tenth to a third of that length. Nonetheless, it’s a factor.
The losses in the antenna depend upon 1) the quality of the grounding (counterpoise) AT THE ANTENNA, and 2) the losses due to the materials and construction of the antenna itself.
The radiating element of an antenna needs something to "push" against. In the case of a dipole antenna (not common in mobile or portable installations) one side (pole) pushes against the other - it is "balanced" - like a teeter-totter. In the case of a monopole (aka whip) antenna, it pushes against a counterpoise, usually chassis ground. A poor ground means that it has little to push against, so it won’t radiate efficiently. A GOOD RF (Radio Frequency) ground for a mobile installation is difficult to obtain, and is generally made up of a massive amount of electrically connected metal, to which the "antenna mount" is connected. The antenna mount for a CB mobile setup is the metal bracket where the coax cable ends and the antenna begins. So, an antenna mount, attached to a mirror, might be an acceptable arrangement PROVIDED that the mirror bracket is made out of metal, and there is a metal-to-metal (electrical) connection to the metal frame of the RV coach. Hopefully, the RV coach is electrically connected to the RV chassis at SEVERAL points. One can do a rough check of this by using an ordinary continuity checker between the mirror bracket at the point where you plan to attach the antenna mount and, say, a (metal) bumper or the frame. This won’t guarantee a good RF ground, but is a good place to start (i.e., necessary, but not, by itself, sufficient).
An idealized monopole (whip) antenna is a straight rod of metal, one-quarter wavelength long, positioned perpendicular to, and with one end near, a perfectly conducting ground plane. So, on the 11 meter Citizens Band (~27 MHz) this would be about 2.75 meters (11/4), or about 108 inches (9 feet) in length - pretty long! Not something you would want to attach to your mirror and drive down the road at 70 mph or under bridges! Especially since it would be best to have it pointing straight up (for reasons that should become clear later). There are things one can do make a resonant whip physically shorter than this, but they all decrease the radiation efficiency and/or increase the losses. One common practice is to spiral-wind the metal rod (wire) around a non-conducting rod, such as fiberglass or plastic. This has the effect of reducing the overall physical length while maintaining resonance. Unfortunately, this also increases the losses. Rule-of-thumb: All other things being equal, longer is better.
Moving on to the spatial distribution of your signal...
As stated earlier, an idealized whip antenna is a straight rod of metal over a ground plane. I should have added "... with no other metal close by". If no other conducting materials are close by, this idealized scenario will produce an omnidirectional signal pattern, i.e., a pattern that is the same in all directions of azimuth. If, however, other conducting objects are nearby (closer that a wavelength, 11 meters or about 36 feet, away), the signal pattern will start to become directional - more signal in some directions and less in others. Larger, closer objects will have more of an effect than smaller, distant ones. E.g., something a foot long and 10' away will not be a significant factor.
This may explain why a short, mirror-mounted, whip may perform better, in spite of the higher losses, than a bumper-mounted, 108" whip, if the bumper where the whip is mounted is six inches away from the metal framework of the coach body. Unpredictable behavior can also occur as the whip swings around in the breeze, as you go around corners and under bridges, getting closer to and further away from the metal as it does so.
And, finally, on to multiple antennas, specifically two whips, each mounted on a mirror bracket. Assuming that 1) both antennas are the same model, 2) the mirrors are symmetrical about the coach centerline (same height above the ground, etc.), 3) the mirror brackets are both properly grounded and 4) the same length (within a couple of inches) of coaxial cable is used from the "splitter" to each of the antennas (proper phasing), the strongest signal should be along the longitudinal axis (front-to-rear) and the weakest signal should be along the lateral axis (side-to-side). Generally speaking, depending upon the distance between the antennas, there may also be additional "nulls" in the radiation pattern of the antennas as one proceeds around the plane of the antennas. However, in the case of a motorhome, where most are about 8' wide, and the antennas are about 9' apart, there should not be ANY nulls, since the antennas are less that a half-wavelength (which is about 18') apart.
So, all other things being equal, I wouldn't expect a pair of mirror-mounted antennas to do much for you. Or, for that matter, against you either. However, attention to other matters, especially proper RF grounding, can make a BIG difference.
YIKES! This went on MUCH longer than I expected. I hope some of it is useful.