They just set aspirational targets, but don’t define any technology for meeting those targets. ITU jumped from 200 kbps for 3G to 1 gigabit for 4G. That was not a useful definition and the market correctly ignored it. Imagine if we had “3G” wired networking defined as 100 mbps and “4G” wired networking aspirationally defined as 10 gigabit. What would you do when the IEEE (the folks who actually make the technology) release 1G Ethernet. Would you insist on calling it 3.5G wired networking?
“4G” is LTE. It’s 4G because it’s what came after HSDPA/EVDO, not because it meets any particular ITU target.
Also, it's disingenuous to describe the jump as from 200kbps to 1Gbps, since HSDPA offering 14Mbps was already being deployed prior to when this particular ITU standard came out.
(https://www.gsma.com/aboutus/gsm-technology indicates deployments of HSDPA started in 2005)
The problem here appears to be not that AT&T is departing from the ITU monikers, or using 5G to refer to LTE Advance Pro, but that it’s selling something as “5G” that’s not faster than “4G.”
The last time AT&T did this the ITU's response was "4G can now also be 3G with 'substantial improvements'"
Source: http://www.itu.int/net/pressoffice/press_releases/2010/48.as...
https://www.itu.int/dms_pubrec/itu-r/rec/m/R-REC-M.1645-0-20...
page 9 defines high / low mobility as follows: "low mobility covers pedestrian speed, and high mobility covers high speed on highways or fast trains (60 km/h to ~250 km/h, or more)." (this is roughly 35 to 150 mph)
We are nowhere near those speeds, even on the best networks.