And they perform like shit. The mirror arms provide no ground,
The bus in the pic is metal, so assuming there is a properly bonded connection between the mirror mounts and the bus' outer shell, there will be plenty of ground plane. Further, with a set of "co-phased" antennas you don't need as much ground plane as a single antenna, they work off of each other to a point... That is not to sat that the antennas won't perform better with more ground plane, but to get "co-phased" antennas to function, you don't have to have as much.
the spacing of the antennas is barely sufficient for co-phasing and it turns your signal into one with a pattern that is from left to right and far less omnidirectional so you could be talking to someone with say a S5, turn a right corner and find you can't hear them anymore.
Which is it, is the proposed antenna setup is barely enough for co-phasing, or is it enough that you will get such a drastic difference that an s5 signal will no longer be heard simply because you made a turn? You can't have both as a result of "co-phasing", they are mutually exclusive results so at best it will be one or the other. So which is it?
Now I'm not a fan of "co-phasing", in my experience in *most* cases a single antenna mounted along the center line between the "co-phased" antennas will generally perform just as well. I say most because I have a model of my Ford Explorer where such antennas in a certain position actually nearly got 2 dB more gain towards the front compared to a single antenna mounted on the roof on the middle line between said dual antennas. This is not enough to make an S5 signal unhearable because of a turn. No other "co-phased" model on said vehicle did that, and it wasn't expected in the first place. It was more a result of antenna placement than weather or not the antennas were "co-phased"...
If you have two "co-phased" antennas at 1/4 wavelength apart, the ARRL antenna books 15'th edition and newer clearly state than you can expect to see 1.1 dB gain over a single identical antenna mounted between them. You can mount them closer together, although the small amount of gain they can provide drops. You can have a set a few inches apart and still call them "co-phased' if you want, even though they will ultimately perform like a single antenna. This gain will also be broadside of the antennas, or if they are mounted on a vehicles mirrors, gain will be towards the front and back of the vehicle, not to the left and right sides of the vehicle. Modeling confirms both of these. Modeling has also demonstrated, at least to me using my Ford Explorer model, that the shape and size of the vehicle, and where the antennas are mounted on said vehicle, may well have more of an effect on the radiation pattern than the antennas themselves, and weather or not they are a single antenna or a "co-phased" set...
I don't know that I would suggest jumping directly into "co-phasing", I'm not a fan of it and never have been. On a vehicle it adds a bit of complexity, and more failure points, all for gains that you will rarely if ever notice... For that bus, if anything, I would say try as many different setups (as in antenna mounting locations) as you can over time and see how they work for you.
The DB