This is certainly strangely phrased. The version I got from another member was that the content of the talk was to be about a proposed feature that may not actually end up in Rust (in my view it likely would not), and having that be the keynote talk would imply to a lot of people that it would end up in Rust and put the language in an uncomfortable position where something other than the feature's merit would be pushing for its inclusion. Moving the talk away from being a centerpiece, in such a case, makes sense, and I have no idea whether this counts as poorly handled or not but saying it shouldn't have happened at all seems a bit naïve to me.