I think there's probably only one way to ultimately test to see whether or not the Duo has a 2x drive in it, and that's to run the Duo (region-mod of course) with an arcade card, running a heavy memory game, and then running the exact same game on an original setup with the arcade card pro and testing the time differences. The usage of the arcade card should bring both drives to their max spin, which is the only time you can accurately test their speeds. Obviously this test wouldn't apply to the Duo-R or Duo-RX, because we're not comparing those...we're talking the US models here. Who knows...they might have actually put a 2x drive into the RX, since it was designed to run ACD games...and there, a 2x drive WOULD help a whole lot....a 4x drive moreso...
Regardless though...doing a speed test on normal 64k or 256k games isn't going to do much, since there's so little data to be transferred. A 2x cdrom has 2 or 3 seconds spinup time before it reaches max spin, and by that time, most all the data's going to be loaded anyways...it would take approximately 3 seconds to fill the entire 256k, and then we can be generous and say we're also going to fill the 64k ADPCM buffer...let's assume we're loading an entire 256k overlay...3 seconds, it's loaded, then reseek to the overlay with the ADPCM data...we're already at max spin, so our transfer rate is 300k/s...it should take less than .2 seconds to load that 64k buffer once the laser is in the right position. Now let's look at the same test in a 1x drive...max spin is achieved in under a second, our data rate is halved, so it's only 150k/s. But wait a second here...we've only got to load 256kb, so we're probably going to load 150kb in the first two seconds, then fill the rest of the buffer in the next second...max three seconds. Then the 64k ADPCM buffer fill...still at max spin now, reseek and load, we're looking at less than .7 seconds. Look mah...our fourth grade math course has probably taught us that we've saved a whole .5 seconds...at best.
This whole test, of course, is not taking into account the speed of the read head, which is instrumental in changing tracks and also affects the seek speed on the same track. We already know that the Duo's CDROM drive has a faster read head...as I stated before, you can tell this from listening to both drives in action...the system 2 drive is quite the slow one, but the Duo's read head zip back and forth pretty fast. On games with heavy track seeking (such as CF2), this makes quite a difference in load times. But the fact that the read head is that much faster will indeed account for the load speed difference...even if the Duo does indeed have a 1x drive.
