As always with statistics one has to be very careful with interpretation. Here's some questions for you to consider: What data was collected? What does a single result and the total result really represent? What do I want to know and how can the data tell me that?
I think you're misreading that list when concluding that a higher percentage really indicates a better drive at DAE. Just look at the raw data:
Drive: PLEXTOR - CD-R PX-230A (63 users): 3505 Accurate Submissions 11 InAccurate Submissions 99.6871 % Percentage Accuracy
Drive: PLEXTOR - CD-R PREMIUM (298 users): 9625 Accurate Submissions 167 InAccurate Submissions 98.2945 % Percentage Accuracy
Drive: PLEXTOR - CD-R PREMIUM2 (85 users): 655 Accurate Submissions 11 InAccurate Submissions 98.3483 % Percentage Accuracy
To begin there is a huge variance in users. Also the ripped media were not the same for each drive (the method tries to make up for that by getting a large number or results, however the variance in users kind of spoils that effort).
Want to see some really wild stuff? Four times exactly the same hardware (with / without RAM support and each with PATA or SATA):
Drive: PIONEER - DVD-RW DVR-112 (51 users): 445 Accurate Submissions 20 InAccurate Submissions 95.6989 % Percentage Accuracy
Drive: PIONEER - DVD-RW DVR-112D (181 users): 4128 Accurate Submissions 220 InAccurate Submissions 94.9402 % Percentage Accuracy
Drive: PIONEER - DVD-RW DVR-212 (100 users): 2820 Accurate Submissions 92 InAccurate Submissions 96.8407 % Percentage Accuracy
Drive: PIONEER - DVD-RW DVR-212D (94 users): 2393 Accurate Submissions 36 InAccurate Submissions 98.5179 % Percentage Accuracy
RAM support shouldn't affect DAE at all and neither should the interface. So why would there be a difference in percentages? It's the different media that where used that we are looking at.
IMHO that list is pretty much useless when it comes to find the best DAE drive. Even by collecting a lot more data I guess it wouldn't become more meaningful because the results would probalby balance out.