I agree that optics itself cuts away certain wavelengths of UV while coatings found in modern lenses will, in addition, absorb some UV. Although it might be significantly lesser with a modern lense, it doesn't mean that there wouldn't be any UV left as even UV filters don't cut out UV completely.
I also do not agree that using UV filters on DSLRs has no additional benefits, judging by your statement. It serves an additional layer of UV screen and helps to reduce purple fringing which is partially cause by UV. It's much like buying a sunblock SPF 30 for more protection over a SPF 15, that is if you are willing to spend more for it.
UV is not known to affect AF, which means that your experience at 5900m (where is that may I ask?) is correct. It is, as discussed earlier, known to degrade the apparent sharpness and affect the colour tonality of your photos.
Not knowing the atmospheric, ground scattering conditions or the transitivity of lenses for UV in any of the cases mentioned by TME and yourself, we cannot draw any conclusion that UV was indeed the culprit.
Furthermore, there are some other possible causes of TME's photos as surfaced in the DPReview thread. I doubt the real cause can be found easily with so much factors that have be taken into account.
