This is how Site Import works. The pages on the live site have the same name but they dynamically display different content based on the values in the URL parameters. Site Import does not have access to any of the server side code so it makes individual pages for each of the different URL parameter configurations that it encounters for a given page.
What you get in the end with Site Import is a static version of the site. Have you tried specifying the directory that the 2nd level pages are in as the staring point for the import? This should let the process start at a sub directory and crawl the site from there. You can also increase the number of pages to import to account for more pages that are in the site.