MediumRare, I don't have exactly the same configuration as you since I have a 166S, not a 163. Also, I don't do much DAE, and when I do it is generally from very clean CDs that either drive reads just fine, even in burst mode. Therefore, I can't give you a specific recommendation other than to say stick with whatever works better for you rather than what some web site claims is theoretically better.
There are a number of reasons EAC support may say one drive is "better" at DAE than the 48246S, and some of those reasons may not apply to you and your setup. Here are a few I can think of:
1. A drive is better because it doesn't cache audio, making it easier (and paradoxically faster) to detect errors by reading the same data twice and looking for differences (data re-read from cache will never be different, since it is physically read just once).
2. A drive is better because it reports C2 errors more accurately (thus, you know when and where errors occur and can take action to correct them).
3. A drive is better because it generates fewer C2 errors when reading problematic discs.
I think that EAC support tends to concentrate on the first two criteria. However, this doesn't necessarily tell the whole story, and a drive could win on points 1 and 2, but still lose out to a drive that does better on point 3.
For example, postulate a problematic disc with many defects. Now, imagine drive A that easily and accurately detects C2 errors with guaranteed 100% accuracy (humor me, please). However, drive A can never manage to read the defective areas of the disc accurately. It knows with perfect certainty where the defects are, but it can't get the correct data.
Meanwhile, imagine drive B that completely falls on its face with respect to points 1 and 2 (let's say it doesn't report C2 errors at all), but due to its superior optics, carriage and servos manages to read that same problematic disc straight through the first and every time with no errors. If there were errors it would have no idea where they occurred or even that they occurred, but it can read without any errors (again, humor me while I take this to the extreme).
Given this scenario, which is the better drive for DAE? Drive A that can tell you exactly where the errors are but can't tell you correct values, or drive B that has no idea where the errors would be if there were any, but can read the disc without any errors?
A rhetorical question of course, but I hope it illustrates some of the difficulties in declaring something "best" (or even better).
In your case, the 163 wins on point 1, I don't know which wins on point 2, and it appears from your test results that the 48246S wins on point 3. Remember, though, that it is possible that although the 48246S is consistently returning the same data, it may be consistently returning the same wrong data. And keep in mind that perhaps your 163 would be the better drive if it were in tip-top shape, but may be having troubles due to age, dirt on the lens or in the mechanisms, etc.
You might try enabling the use of C2 pointers on the 163 and see if that makes any difference. I don't know how accurate the C2 reporting of the 163 is and, thus, whether or not this is recommended. Still, you might want to try just for kicks and to match your settings for the 48246S, even though I doubt it will help given the magnitude of the problems your 163 is experiencing.
Also, when comparing data extracted from different drives, you need to adjust the offsets or the data will never match correctly. I believe the 163 has an EAC offset of +594 and the 48246S of +6:
In the end, the results that you actually obtain with your actual drives are what really count. So don't feel that you are betraying EAC by using your 48246S instead of your 163. I say if it sounds good, it is good!