I think we're getting some cross-threading here - but those high frequencies are really important to speech quality. I don't know why people always focus on vowels and sonorants to the exclusion of everything else when this issue comes up. Stops and especially fricatives have significant spectral energy way up into the 10khz range.
And is all completely pointless in the context of this thread, where we're talking about SSB - the lowest quality voice transmission known to man.
There's no need for high frequencies, or high bandwidth - if you want that then meet in person and talk. But even with the low quality of USB perfectly acceptable speech is the result (at least on a properly designed system).
Attempting to do high bandwidth audio would not only be illegal (breaking the licence conditions), it would also limit your range to down the street, rather than across the Atlantic ocean.