I can answer some.
The number and distribution of PIFs should be almost identical, +-1 or 2 in individual values, +- maybe 10% in the total number. PIEs will vary a little more from scan to scan (I've seen >30% differences) due to different disc alignment, temperature, etc., but the general shapes, slopes, and other PIE patterns should be similar.
Much more variation. Large defects should be recognizable in both scans, otherwise there are too many differences between optical and ECC (error-correcting) performance of two different drives to make any meaningful comparisons.
Yes, the varying speed is normal CAV (constant angular velocity) reading. The spindown around 130 MB is always there, in all 1620s.
There is a scanning speed selector in the quality test window of CD-DVD Speed that allows you to scan (on the 1620) at 12x and 16x CAV, but 12x and 16x scanning is problematic due to multiple spindowns and their effects on the results.
Ritek media always had high PIE levels, nothing unusual there.
You can't miss a black dot on a quality scan. Here are three scans from an experiment I just did.
The first scan is a freshly burned disc.
The second scan is the same disc with three black dots, about 0.5 mm, 1 mm, and 1.5 mm in diameter, made with a black Sharpie pen using quick-drying, alcohol-based ink. You can easily tell how they were arranged by looking at the scan. The DVD error-correction system, called RSPC (Reed-Solomon product code) can compensate for up to 7.1 mm of missing or damaged track data, so it has no problem reading through my dots.
The third scan is the same disc after I wiped the dots off with a tissue soaked in alcohol. Compare scans 1 and 3 to see the normal variations between scans of the same disc in the same drive.
(Scans 2 and 3 are partial scans, manually stopped after 0.5 GB.)