Since then, we got HandleURL which enables serving a virtual directory structure. In terms of code, the command Request.Print within HandleURL enables serving standard HTML pages content stored for instance in constants.
See http://documentation.xojo.com/index.php/WebApplication.HandleURL
The trick is then to serve the regular application when a regular browser hits virtual directories, because that is the way search engines work. If they find a particular content, say, in page 3, your app must be conceived for the user to land on that page and not the main page.
If I may, a web app is not a web site. Very often, the textual content of an app is extremely limited as compared to a regular web site. So even if one can serve pages to the robot, the content may be too poor to get a good ranking.
Note that Ralph’s app https://www.aerorasor.com/ is a catalog, which lends itself much more to search engine indexing.