What happens is that you select in the inner for-next loop all the first thumbnails of each media:group node. Instead you only want the first media:thumbnail of the current item in de outer for-next loop.
You can solve this by changing the line with 'Set xmlC1Nodfes =' to:
Set xmlC1Nodes = xmlEntryNode.selectNodes("media:group/media:thumbnail[1]/@url")
Two things are changed in this line. In the posted code you used xmlDoc.selectNodes, thus referencing the whole XML document. Changing this into the xmlEntryNode.selectNodes, will give you a reference to the current entry.
This will not solve the problem, beause off the trailing '//' in your XPath query (searching from the rootnode of the document). Remove these to start your XPath query from the xmlEntryNode.