If you honestly think PG rejected people based on their idea, I think you're wrong. That's probably the last criteria he used. So, work on the idea. I would reject 90% of submissions, too, if I was running something like this.

For example, with Google, Google AdSense was the glue that made everything work (profitability, great product, many users, good reputation, ability to have free food.) You might say their demo and pagerank was actually important, but they couldn't even sell that for $1 million dollars. Heck, even right now, nobody would buy their original technology as Yahoo! and MSN must have already surpassed the original PageRank in quality.