If your URI stems are dynamically generated, are there other sources of information in the request or response that we can use to track a page identity?
For example, Siebel uses the same URI stem (start.swe) for all its "screens" but the URI Query parameters can give you enough information to designate something as a page and its name.
Do your pages carry any information like these in the POST parameters, Cookies, HTTP Headers or HTTP content? Using system fields and primitive custom fields you can extract information out of the URL, HTTP headers or HTTP body to assign a name to your pages when the URI stem is not useful.
Right now, what does the Session timeline or Session browser show when you drill down a single session to a list of pages?
EUEM is able to extract content out of the Dynamic URL Taking advantage of Primitive Custom fields can allow you to get the information your want our of these dynamic URLs.