Oxford University Press has published A Dictionary of Modern English Usage: The Classic First Edition. Nothing especially notable in that, except for bibliophiles and usage scholars. But what sets this publication apart is David Crystal's introduction to the volume, an assessment of Fowler's entries.
My copy has not yet arrived, but I've read notices of the volume in several places, in particular an appreciation of Crystal's commentary by Language Hat. A perceptive quote from Crystal:
The problem in reading Fowler is that one never knows which way he is going to vote. Is he going to allow a usage because it is widespread, or is he going to condemn it for the same reason? … The impression the entries give is that Fowler considers to be idiomatic what he himself uses. Usages he does not like are given such labels as 'ugly' (e.g. at historicity) or even 'evil' (e.g. at respectively).
Language Hat follows the theme:
[Crystal] continues with a good deal of acute analysis of Fowler's choices, prejudices, and insights, presenting some striking examples of contradictions: Fowler sensibly rejects letting etymology define meaning, then turns around and expresses "strong support for the maintenance of earlier meanings of a word, such as at aggravate, transpire, and meticulous (a 'wicked word')."
It's not just Fowler, of course. Observers of the advice literature (here on Language Log and elsewhere) have noted that usage advice is so often an expression of the adviser's personal (and not infrequently idiosyncratic) taste. For a recent example, see the discussion here of David Foster Wallace's recommendations about grammar and usage.