The math works out very, very differently for traffic conditions than it does for store availability. With traffic conditions, with a full road of traffic, if even only a couple percent of people have Waze, and then even only a couple percent of
those report an accident, for example, you can still be fairly confident the accident occurred.
With stores that have thousands of products, and the app is only useful based on it's accuracy for each individual product, you need many, many more people reporting because their reports are spread over so many products.
TBH, I don't think this idea is workable, mainly because for the most in-demand items where you want a service like this, the inventory changes by the minute. I mean, whole aisles of toilet paper can easily be cleared out in 10 minutes even when purchases are limited to one package per customer. It takes me 10 mins to drive to the nearest grocery store.