To be the devil's advocate: Breaking a result up into little pieces to increase your paper count ("salami-slicing") is frowned upon.

Of course this is not what Terry Tao tried to do, but it was functionally indistinguishable from it to the reviewers/editors.