Imagine you're doing a sector-by-sector scan of a massive, corrupted hard drive taken from a crime scene, trying to find out who the owner of the drive was. It makes no sense to do a rigorous analysis of the first possible file fragment that turns up - the odds are, it's just some random data file that could be found in any system. How many people have memories of Denver, yet have never been there in person? How many people have been to Denver, with no records remaining of their visit?
There is no reason to believe this man ever had anything to do with Denver, especially given the reliability of his "memories" thus far. How could someone be named "benjaman", yet not turn up in any written records, anywhere?
I have very vivid, intense memories of Las Vegas, that are tied to emotional experiences I have not had elsewhere. I spent only a week there, and paid for everything in cash. Would it be useful to start the search there?
The only reason the method I suggested would work, is because it relies entirely on weighted statistical analysis of what region-specific symbols this man has been exposed to. You can't fake strong recognition - it reflects changes to neural structure over long periods of time, and has a very low probability of false positives.