GREat (or not)

My recent post at the Valve has spawned an interesting subthread in the comments about the GREs.   Jonathan Goodwin argues that "the GRE subject test is directly correlated with a traditionalist curriculum," and further suggests that "you can generally guess that the average undergraduate from a university with a doctoral program in English is going to have had a less traditional curriculum and probably would not do as well on the test..."  This latter point was certainly true of UCI in the early 90s, when I took the subject exam; a couple of my professors warned me that because the department's requirements for historical coverage were fairly open-ended, English majors often stubbed their toes on the GRE.  Since I had actually pursued what Jonathan calls a "traditionalist" approach--Chaucer, Shakespeare, Milton, the works--I did well enough on the exam to flap one normally unflappable professor ("You got a what on the where?!").  But even so, the exam struck me as a fairly odd bird.  In the 90s, at least, it was really a test on the various and sundry contents of the Norton anthologies, and good test-taking skills frequently seemed more important than actually knowing all that much about the material.  (I remember one i.d. question that repeatedly incorporated the play's title into the plot summary.  Um, golly, I wonder what the right answer could be...)  Did it accurately test what I knew, or just what I happened to know of?