It made sense to me. It's a great idea but implementation is going to be incredibly challenging. It's only doomed if the founders don't have the skills to actually implement a better recommendation everything.
I think an uber-short pitch here is "A better recommendation engine for everything" which of course only makes sense if you know what a recommendation engine is.
Even if you built the most perfect possible recommendation engine, it still wouldn't matter unless you were able to build a large enough user base who was actually using it.
Re: ability to get recommendation engines working better, see the Netflix $1M competition as an example of just how hard it is.
I think an uber-short pitch here is "A better recommendation engine for everything" which of course only makes sense if you know what a recommendation engine is.