Correlate your source for the Window ID.
There are easier ways of handling this issue however. When you run in HTML mode, versus URL mode, the page is brought into memory and parsed for the resources to request downstream. You can effectively comment out the extra resources noted in the script (EXTRARES) and run your HTML mode script, have the virtual user parse the page and load the extra resources based upon both the browser emulation settings as well as the cache headers for the resources.
Don't take my word for it. Comment out the extra resources and then run with full logs for a couple of pages. You should see in your logs the evidence of the page being parsed and the resource files being requested as needed.