Weewinkywilly, there are some nice HF transmission tools that chiefly work via NVIS, Near Vertical Incidence Skywave. This depends on common properties of the ionosphere over the HF spectrum. It can be reduced to virtually a telephone technology using ALE, Automatic Link Establishment. It is claimed the military squeezes as much as ISDN level bandwidths out of HF, 128kbps. Link establishment time is small. And for data transmissions of low data rate aircraft information these might do well.
Again, two problems exist, frequency allocations and polar regions.
Frequency allocations can be squeezed out, perhaps. While HF broadcasting is in severe decline squeezing out more bands for amateur radio has proven very difficult. Were these inactive or no longer needed HF broadcasting frequencies no longer needed obtaining new ham bands should, logically, be easy. So I suspect this will be a serious problem even for aircraft safety issues. A lot of frequencies will be needed and significant weight additions for the new transceivers, antennas, and computers will exist. These can all be dealt with.
Polar regions remain the tough spot. I am not well versed in the ionosphere's configuration at the magnetic poles. I'm not sure it has been studied. It WILL be vastly different due to the interactions with the Earth's magnetic fields. And we have auroras in the polar regions. Auroras shred and spread signals. Most transmissions, AM, FM, or CW sound more like they've been fed through something for audio like shredding is to junk credit card offers. SSB works, in so far as it is "understandable", amazingly so compared to CW. AM and FM are generally unintelligible unless the power is very high. Since SSB can get through that suggests that somehow data should be able to get through at about the bit rate for vocoded speech, 100 to 200 bps, using brute force techniques. Maybe an off the shelf technology to deal with it exists. I'd expect it still would be classified if it does. How many wars do we fight in the polar regions?
Maybe this can be done. But the proposal needs more polish performed by people with more current experience with recently unclassified systems than I have at this moment.