...my first app, which allows you to enter your 20 favourite recipes. It will then choose a week’s worth of dinners at random, and compile a combined shopping list for you, and let you tick off the items as you purchase them. Now the market has decided that this idea isn’t worth paying $1 for. [...] I can only conclude – the idea sucked.

This actually doesn't sound like a bad idea at all, and would definitely be something I'd get if I stopped working late and started cooking more. Anyhow, a good idea and a strong implementation do not guarantee success.

I'd guess one reason it didn't sell well is because nobody thinks to themselves "Oh, I need an app that randomly picks meals I already know how to make and gives me a grocery list." Rather, this is an app that people will stumble upon and decide whether they think it is useful or not. Perhaps the missing ingredient for this app's success was neither idea nor execution, but rather marketing and a bit of well-timed publicity.

Edit: Found the app (http://itunes.apple.com/us/app/random-chef/id438618418?mt=8&...), and I would double down on my wager that it is a marketing/presentation problem: those first 2 screens definitely do not look very appealing. I'd scratch the paper texture--or if you really do want a paper texture, use something that you'd actually write a grocery list on, and not an ancient manuscript. The multiple different fonts throughout the app don't help with the readability or the presentation. Anyhow, the idea still sounds good, but how do you expect people to think it is well implemented when the first two pictures they see look like a mess!