I pitched this idea to YC last year, basically.
Mine also had "GPS on phone to monitor person's level of physical activity to help tailor diet to their dietary needs"
We even worked on putting this together so that the algorithm could work automatically for a family who had different dietary requirements.
The problem we found is that it doesn't work for the consumer market. We focussed too much on the idea and not enough on consumer behaviour, which if we had done our market research would have shown this to be not so good an idea.
We found that the dietary consumer market is focussed on selling you early into things you wont use, because the people that generally need these products have low motivation.
Those people who buy dietary books, supplements, exercise equipment etc generally purchase them and might use them for a month or two before essentially "giving up" at best. That's why most businesses targetting that large market are focussed on getting your money early, because relying on longterm revenue is a bad idea.
Weight Watchers and Jenny Craig use a similar model, although the diet plans they use are actually loss leaders so they can sell you their other products, which is where they actually make their money. The diet plans in the long run is where people stop going, but they've already extracted value out of you by purchasing their material.