Well I'll talk about one thing, personalization and the restaurant industry. While it's great to be able to give customers exactly what they want, we also have to be aware that 90% of the people out there are 90% the same. They wear the same jeans, use the same browsers, they shop at the same stores, buy the same toilet paper, eat the same burgers... etc. Most astute restauranteurs have already honed in on this. Note the number of cookie cutter restaurants out there serving different combinations and permutations of the same darn thing. Boring as heck... but people still like their seasoned curly fries and deep fried cheese sticks.
Personally, I find it all too weird. I used to do the same thing 15 years a go when I was in college IRC'n over a Coke and a microwave burrito. But that was cheap, I had no money, and I had three term papers in the background. Nowadays I just carry around my ipaq. (that's how I replied to your post the first time.)
Anyway I applaud you for your bravery, but I think your concept needs more testing. Usually you don't really come out of focus groups with a fully developed product. You come up with a prototype and present it for either in-home tests or in your case a test market. The reason is statistically you're trying to find an intersection between two sets with enough members such that you can sustain a business. Of course! That's how niche markets are created! The problem is that it's not a priori provable. (Those familiar with the axiom of choice, also known as axiom of multiplicity will get a chuckle. Finally, someone who actually understands that this conundrum has real life application. There's hope for geeks after all)
In otherwords, in simple English, you will not know unless you test it.