Will try to explain the casus as clearly as possible, so please bear with me ;) The sitemaps I generated for the GED on pre-4.4.2 versions (I'm using PGV since version 3.1) were always constructed just fine by PGV. As of v4.2.2, however they contain duplicates and incorrect entries.
My site is built on 6 GED files. One is the main GED, containing most of the individuals (about 10.000). The other GEDs don't link to the main tree.
Now, for the main tree, the individual URLs in the sitemap look like:
In the secondary sitemap GED, I'll find:
The PID=5161 is invalid for the secondary GED file, as it doesn't contain this PID at all.
As a result, the 6 sitemap files for my GEDs will all be about the same size. With older versions this used to be 1 large sitemap file, and 5 significantly smaller sitemap files. This was my initial indication that 'something' was wrong...
Feel free to contact me for additional details.