Here at Wilson F/X we do have some experience with the 900hp units in a moving vehicle. We were testing the range while I was driving away from my partner. He had the pad-box stationary and I was driving with the controller powered and the bank/pad selected. the units maintained constant contact till I got to a low spot in the road and lost line of sight. I reacquired full signal after another 100 yards or so. We were in constant cell phone conversation too. We had full contact out past 2 miles here in the mid-west. We eventually lost line of sight due to too many hills, but it gave us the info we needed.
I did the same on my own out at Black Rock at the last LDRS that was there, and kept a working signal out to 1.1 miles and fired a "significant" ejection charge just to make sure I had not only a return signal but a working connection. As ought to be obvious, this was all on the ground, but it was also full communications both ways from controller to pad-box and from pad-box to controller. In very much layman's terms the two-way communications looks like this: full normal automatic call from the controller for selected specific bank/pad continuity, response from the specific bank/pad-box of continuity status. Then (after deselect of the pad) a specific order from the controller for a report of the voltage at the pad-box and a report of voltage from the pad-box. Followed by (reselect of the pad) stand-by continuity signal and standby continuity report back from the pad-box, command from the controller to fire, response of pad-box by firing that specific pad-box, (release of the "fire" button) followed by automatic report of "after-fire" continuity by the pad-box, and reception of the report of after-fire continuity from the pad at the controller. Followed by deselection of the specific bank/pad switches which disarms that bank of pads.
Now whatever else you might think, that's a lot of encrypted information to be passing back and forth between the controller and the pad-box, but its pretty much standard except for the voltage reading which is only done if you specifically request it. So, XBee's can handle a great deal of info in a very short amount of time.
With a custom WFX hybrid control system we've sent reports of the "pre-fill" weight of the rocket and the "post-fill" weight of the rocket in order to determine the amount of NO2 that is actually loaded into the rocket. And with this same system we can also report the temperature of the gas at the motor "vent" to also determine when the rocket's NO2 tank is full because when the tank is full it starts venting liquid NO2 which is a bit colder than merely venting NO2 gas. We did this last one because in some environmental conditions its almost impossible to "see" the venting of the NO2 from a full on-board NO2 tank.
So, like I said, XBee units can transmit a great deal of information back and forth, even encrypted.
I am not making any claims about using XBee units in flight. We haven't done that. At least not yet. LOL!
Brad