"The psychoacoustic model provides for high quality lossy signal compression by describing which parts of a given digital audio signal can be removed (or aggressively compressed) safely—that is, without significant losses in the (consciously) perceived quality of the sound."

Significant and safely are the two words you should be reading there. It's not proven 100% of the original lossless file as its literally lossy and getting it right 'most of the time' as their model does, isn't the same as getting it 'always right'. So lossless is still obviously preferred.

I mean you're really taking the whole work out of context as it's designed to lose less of the recording than the data shrink would suggest, not provide a perfect match for lossless.

Not to mention the factors of the dac and amp portion of a tws, few actually measure 'perfectly' despite sonically transparent dacs and amps (such as topping) being affordable both as portables and desktop set ups. So wireless has a ways to go, that said by someone with both wireless and wired set ups.