I think in any area of knowledge it is important to be able to distinguish the difference between a humble but knowledgable person expressing his views in a genuine, self effacing way, from those of an ignorant person, with a lack of confidence, being bullish about his knowledge.

With regard to distortion increasing dynamics, we can take a simple model and deduce the effects of distortion on dynamics.

If we take the example of a signal and introduce 1% distortion, a rather high level, (-40dB), the new signal level becomes 1.01/1 greater in magnitude. It is evident that this increase in level is very low in dB terms and probably undetectable to all.
(I do not have a charged up logs calculator to hand, but fro memory even 1.1/1 gives only about a 1dB level increase).

WRT clipping, of course that will produce odious distortion, well understood and documented, though on the HUG site there are claims that the distortion produced by valves under clipping conditions is mellifluous to some.