Ted Goranson has another "About this Particular Outliner" article out: a detailed comparison of two of the major outlining programs for Mac: OmniOutliner and TAO. The whole ATPO series is fascinating, but the latest article strikes me as rather unique in the field of writing about software. It is not a review in the usual sense. It compares features, discusses strengths and weaknesses, offers screen-shots, but Goranson conspicuously avoids boiling all this down into a "top pick" or "buying advice." Instead, he makes it clear that the two programs are really geared toward different work styles and project needs. He explores how seemingly small matters of interface design can subtly push the user in different directions. Instead of telling the reader what to use, Goranson is really teaching the reader how to make that determination for him- or herself.
All this stands in contrast to recent Macworld "roundup" of web browsers. It's not that I necessarily disapprove of their Editors' Choice in favor of Firefox (although, my choice is different), but in making this choice they embrace a features-for-features'-sake attitude in which individual users' needs are secondary to some absolute notion of "power." For example, Camino's lack of built-in RSS recognition is seen as a crippling flaw, without considering that a user might have some other, perfectly workable method for reading feeds (e.g., NetNewsWire). This is the kind of thinking that leads to bloatware that does everything, but nothing very well.
Quibbles with their judgment aside, the Macworld article is clearly a review, and there is certainly a place for software reviews (just as there are for book, record and movie reviews). What Goranson does might be more properly called software criticism, in the same sense that we have literary, music, or film criticism, and I think there has been far too little of this sort of writing. This makes his ATPO articles all the more significant: not only are they informative, they are setting the standard for en entire genre.