Its a topic concerning many people these days!
I have to say that despite all the comments about "statistically insignificant sample" I actually think the newegg numbers are pretty good if you can adjust for the packing policies. That is, when did they start packing disks poorly, did they stop at some point due to complaints?
Regarding the main question however, if you look at popular hard drives e.g. Seagate barracudas you will find at least a few hundred reviews for each and some will have satisfaction ratings in the 80-90% range (4 & 5 stars), while others in 40% range. Invariably the ones with higher ratings are older drives with lower capacities.Compare the same drive at 750, 1TB, 1.5TB and 2TB and see if the pattern doesn't hold true. Not scientific, but to me that says something.
Granted there's a lot of satisfied people who don't post reviews but calibrating for those in reviewing the data is trivial. As to the sample size, I confess, being a statistician, this is not how I'd construct a survey, but as a shopper, this is certainly better than nothing.
I'm thinking about going with samsung ecogreens - the ratings are higher but on top of that samsung apparently insist that customers do read/write tests on their new drives for several hours with a utility from their website. No idea if this is the same as the tests done by seagate seatools or WD data lifeguard tools, but it makes me feel more comfortable. Anyway they're going into a raid 5 and are on the HCL, so that's the most important thing. Oh, and backups of course!