I've stumbled on blogs describing new table of contents feeds direct from publishers. Aside from wondering what's taking them so long, I've started wondering why publishers don't better aggregate their own data, leaving that to other parties?

Why wouldn't a publisher with a few dozen titles provide subject-based feeds across all their own journals? Or, for publishers with many titles, offer author-specific or institution-specific feeds? (Aggregators sometimes offer the former; I don't think I've seen the latter anywhere.) While a prolific author may only have a couple articles a year, if you're interested in the same research area as scholar Waldo McGillicudy, you probably know his name and would want an easy way to be notified -- pre-publication, even -- when something new is coming out.

It would also be interesting to see institution-specific feeds. Everything that comes from a faculty member at a particular research university or (for large institutions) department.