Henry Adams' most recent foray into the foibles of his professors included, among other things, a classic example of the "shame the authority" trope. (Incidentally, there has to be an actual name for this, since I see it so often in controversial literature: the student/child/whatever asks an "innocent" question of the professor/priest/whatever, only to leave the authority figure completely flummoxed. Student/child/whatever triumphs; bonus points if the two are on different political, religious, or cultural sides.) "Dr. Ethos" recommends spending no more than "10 minutes" grading each comp paper. Cue Henry Adams, giving in to "the imp of the perverse": "'I raised my hand and said politely, "It's good to know that I need to grade six papers per hour, but right now I can handle only four. Could you and the other professors give us tips on how you reach the six-per-hour rate?'" Needless to say, Dr. Ethos retreats: "'Dr. Ethos looked down at her paper. "You just have to learn to pace yourself," she said. "Let's move on.'"
Impishness aside, Adams doesn't really address the question of how much commentary really assists the students--and to what extent "commenting" becomes its own performance. As some of the commenters point out, spending an hour per paper doesn't necessarily help anyone, even if it makes the instructor feel like a Virtuous Soul. On the one hand, I've heard students complain that they get papers back with no comments on them at all; on the other hand, like every other professor on the continent, the planet, and quite possibly the known universe, I've frequently found that the students don't necessarily read comments once they get them. (Student hands in draft; instructor makes various and sundry corrections on the draft; student hands in revision; instructor can't help noticing that the errors are still there.) When you're faced with a particularly hapless paper, it's difficult to know what to do. Mark every error? (I've done that, leaving the student with more comments than paper.) Say "I stopped commenting here"? (I've also done that, but it feels cruel.) Stop commenting without calling attention to the fact? (Then the student wonders if you also stopped reading.) Trot out "See me immediately"? (This generally gets us somewhere, although I often have to hold back the grade until the student shows up.)
I long ago gave up the stopwatch approach. Instead, my practice is as follows:
1) No more than seven papers at a time.
2) More and more, I find myself marking basic errors once, then telling the student that it's his or her responsibility to find the rest.
3) I summarize the overall gist of my comments at the end of the paper, right before I give the student a grade. (I usually try to start off the summary by saying something positive about the argumentation.)
4) After every batch of papers, I compile examples of the most frequent errors, append a cutesy title ("Law and Order: Grammatical Intent"), and discuss them with the class. At the very least, even if the students fail to read the comments, they'll still hear me talking about misplaced modifiers and the like.