I don't look at the technical details concerning what a drive should or should not be able to do. Clearly the 1648 "can" use CDSpeed to scan a disc. What I have done is compare a number of different discs by scanning them both in a 1648 and then in a Liteon to see if there was a correlation between the results. The CDSpeed author agrees that the results on the AOpen are too high but that you should still be able to use the results to see a pattern. Others feel that you should be able to take the results and divide them by 4 or 8 and see meaningful scans.
What I found was that there was no reliable correlation between the scans in the two drives. Some discs looked too good on the AOpen and some were too bad and some were sort of OK. By the way, there was a clear relationship between Kprobe and CDSpeed for the Liteon.
I am making no statements as to the meaningfulness of scans in general; I find them very useful and a very good measure of burn quality. It is also possible that there is only a problem with my drive, but I see no reason to believe this. I am just saying that scanning with ROM drives in general is unreliable and not meaningful. This makes it useless unless you check it against a burner. At that point ROM scanning is a waste of time.
As to its reading ability, I am a strong advocate. German testing has spoken of superior error correction and I, as well as others, have seen that it has very fast reading ability. I have only found one very bad quality disc I couldn't read. In all other cases, the drive speed just slowed down to get through the errors and then returned to speed after it had passed. If it was unavailable, I would use Liteon, then Toshiba, then Benq.