There are three angles on cooking that I don't find in standard recipe books:
1. The structure of cooking. A lot of recipes share common patterns, but in recipe books they are presented as individual recipes rather than as modular building blocks and techniques.
2. A scientific approach to cooking. By this I do not mean a description of the chemistry, but rather a more experimental approach and a focus on verifying claims based on taste testing and boiling down recipes to their essentials. There is a lot of stuff around food that to me sound implausible or unnecessary. For example the idea that expensive wine tastes significantly better, or that kosher salt tastes differently, or that searing a steak seals in the juices are as far as I know all myths. There are tons of smaller examples, like recipes that call for adding tomato paste, canned tomatoes, and water, or recipes that call for adding the same spice at several different points in the process. If that is experimentally shown to make a difference, then sure, but otherwise keep it simple. On the other hand there are some things that do clearly work in my experience. Increasing the PH of the cooking water keeps vegetables, potatoes, rice more firm, lowering the PH of cooking water turns them into mush. Grinding spices fresh and combining tastes much better than a pre mixed garam masala. Some herbs are ok dried (e.g. thyme, mint), but some are worthless when dried (e.g. basil, parsley). There is a lot of knowledge like that which you could learn by running your own experiments or from a lot of experience, but that takes far too much time for the average person. It would be great if somebody ran the experiments, and for each recipe tried several different ways of cooking it, and determined the best way to cook it with a taste test panel.
3. Quality over quantity & originality. Many recipe books contain a lot of recipes around a specific cuisine or even sub part of a cuisine (e.g. french soups). Many books try to have original recipes rather than classics. I don't care about any of that. Given that I have a limited lifetime, I don't want a book with 50 moderate French soups and 2 great French soups, or a book with 50 moderate Indian curries with an original twist by the author and 3 great Indian curries. Given that I can easily access millions of recipes, the value of a book is not in the recipes which it contains but in the recipes which it does not contain. Please only include the truly great recipes. This also ties in to point #1. This page for example: http://www.jamieoliver.com/recipes/recipe/easy-homemade-curr... It lists 5 curry pastes as 5 different recipes, but 90% of the ingredients are the same. So give a basic recipe, and list the variations. Secondly, it suggests toasting the spices, which in my experience does enhance the flavor of some of the spices, but it destroys the flavor of others. This ties into point #2. Run an experiment: toast each spice individually, then for each spice dissolve the toasted and untoasted version separately in water and in oil, and then taste the difference. I could run that experiment myself, but it would be handy if I didn't have to.