lemon market of terrible management

That is a great way of describing it.

Ah, but from the perspective of a manager who is looking only at the feature lists, what is gained by that rewrite to some other language? The after feature list looks just like the before feature list!

These represent different audiences: the manager looking only at feature lists objects to rewriting the code at all, while the programmer considering a rewrite to another language is seeking to make "this" rewrite "the last one" or at least to minimize the chances of having to do this all over again. The typical PHP programmer only knows PHP — programmers who learn other languages tend to move to those other languages. The typical Perl programmer can be expected to already know other languages but has stayed with Perl because of its nice features.

There aren't very many people arguing seriously that backwards compatibility should be broken willy-nilly, but there seem to be several people arguing that other people are arguing that.

The problem is that it was reported that the individual leading the push for Perl 7 advocated breaking backwards compatibility not for any of the reasons you listed but for purely stylistic reasons. Removing features from the TIMTOWTDI language because you do not like the syntax is not an acceptable precedent to set because then the next guy has different tastes, so removes other features, and the next-next guy does not like something else, and eventually you have a Fahrenheit 451 language (the mass-banning of books in that story began with a little censorship here and a little there, all from different groups with their own proverbial axes to grind) that makes PHP look good by comparison. Every feature removed makes Perl a little poorer. Sometimes there can be improvements to the interpreter that are worth that cost, making Perl richer somewhere else, but citing PBP does not support such a claim.