If you respect the discipline/art/science enough, then you value truth over the ego of some lesser being.
People who dislike the conflict in these kinds of interactions are also inclined to dislike conflict on the editorial page, preferring instead bland "objective" writing.
The thing is, nothing is objective and the best way for the audience to view the truth is by observing the conflict and drawing its own conclusion.
And yeah, Yehuda is one of the best programmers most of us would ever sit in a room with... and thus he would find improvements to nearly anyone's code. To view this as a bad thing and to prefer polite ignorance doesn't seem like the best approach.
There may have also been benefits to the presenter's approach (maybe simplicity?) If you look at Yehuda's refactoring of Rails, there is tons of insight but also significantly more abstraction, which in some cases makes things less obvious. So the presenter should have been able to defend his/her approach at least on some basis.