I wouldn't count on the service being prohibitively expensive. There's a reason the rules are being relaxed, and it isn't because the FCC cares about public demand. Carriers and airlines want additional revenue streams, and those only work if people use the service.
I'm not sure how easy it is to separately filter text and voice on cellular networks, but cellular data service would seem somewhat redundant given the presence of inflight WiFi. Most of the folks I text seem to have iPhones, and I can text to those people now over WiFi. With a majority of cell phone plans offering free texting, I'm not sure that there would be enough of a revenue stream from that feature alone, but perhaps the somewhat annoying trend toward texting instead of talking would mitigate the number of folks who would actually use a voice feature. In any case, I suspect that the industry groups have data that suggests how people would use the service, and how viable it would be to deploy.
I'd also say there's a huge difference in the experience between being on a bus, subway, or train, and being crammed into coach on an airplane for hours on end. The ability to talk on a cell phone isn't an enhancement to any of those environments, and the airplane is certainly the most intimate with the fewest number of options for relief.
JKG