The present marketing spiel is an attempt to alter most people's memories. Based on some of the comments here, I think it may be working.
I still have a vivid recollection of one of the very worst speaker auditioning experiences I've ever had, back around twenty years ago. I can't recall the model, but it was a tall floor-standing speaker, and not their flagship, but still an expensive speaker. I recall being immediately taken aback by just how exaggerated the upper treble was. It was so overt and so unpleasant that I could not keep from sharing my thoughts with the salesperson. It seemed to me that this had to have been done intentionally, and for the obvious reason. It has long been common knowledge that consumers listening to speakers in retail showrooms are impressed by speakers with an unnatural emphasis in upper treble, and are more likely to buy that kind of speaker than a speaker with a flat response.
Exaggerated upper treble is typically associated with aluminum tweeters, however this doesn't always happen with aluminum tweeters, and it sometimes happens with non-metal drivers. Sometimes the true blame is cone breakup of an aluminum midrange, and the tweeter gets the blame. If a manufacturer wants the speaker to have a pronounced emphasis in upper treble, the manufacturer will most likely prefer to use an aluminum dome in the expectation that it will exaggerate upper treble. Manufacturers who do this probably bear the lion's share of blame for the bad rap that aluminum tweeters have.
In the case of that Paradigm floor-standing speaker that I auditioned some twenty years ago, it was definitely the tweeter (not breakup of the midrange), and it was simply a matter of the tweeter having a fairly narrow peak in its response, in the vicinity of 8 kHz as I recall. If aluminum is to be blamed, it is necessary to positively identify breakup of the dome as the culprit. With most metal domes, the breakup occurs higher in frequency than 10 kHz, where most of us wouldn't hear it. And it is easy to identify the breakup range in the response curve, because of the sharp, closely spaced peaks and dips. When we hear an exaggeration of upper treble, it is usually a bit lower in frequency, typically in the vicinity of 6 kHz to 8 kHz. And while there must obviously be a resonance of some sort that is to blame for this peak in sensitivity, the peak in response in this frequency range isn't typically due to the breakup of the aluminum dome. (And it isn't the primary resonance of the driver, because that occurs much lower in frequency, well down into the midrange in most cases.)