I have of course observed dozens of cookery and catering lessons in a range of settings, from exclusive private colleges to further education colleges. Most of them were based on teaching a repertoire of recipes. Granted that the recipes were chosen to illustrate and develop the standard skills and techniques, but the recipes were in the foreground.
The argument for attending principally to techniques rather than recipes is that mastery of frying, or baking, is open-ended and equips one to do anything in the kitchen, beyond discretely-specified dishes.
I really was looking for something else when I picked up Berger and Luckmann (1967) and noted their reference to "recipe knowledge":
"Since everyday life is dominated by the pragmatic motive, recipe knowledge, that is, knowledge limited to pragmatic competence in routine performances, occupies a prominent place in the social stock of knowledge. [...] [A] large part of the social stock of knowledge consists of recipes for the mastery of routine problems. Typically, I have little interest in going beyond this pragmatically necessary knowledge as long as the problems can indeed be mastered thereby." (pp.56-57)The argument is of course that recipe knowledge is convenient and necessary, but inherently limited by its taken-for-grantedness and "typifications" (or use of stereotypes). If my knowledge of cookery consists solely of discrete packages of ingredients and instructions, I will be severely limited in my ability to be original and generate new dishes--or indeed to improve the ones I already know.
(In reality, it is worse than that; I once asked a catering lecturer what was the biggest challenge or difficulty he faced in teaching. He did not have to think twice, "Getting the students to taste what they have made," he said, "they don't even understand why they need to eat it.")
Of course, "recipe knowledge" (the term originated with Alfred Schutz, and includes skills as well as conceptual knowledge) is a metaphor and not confined to culinary skills--it is everywhere. The "communicative" approach to language learning--handling typical social scenarios in the hotel, restaurant, asking directions, etc. is very effective as far as it goes, but it is effectively capped. To engage in a free-flowing conversation you need to look under the bonnet, as it were, and understand the grammar and structure and how they work, and that is far more challenging.
It has been my impression over my 40+ year career on the edges of vocational and professional education that recipe knowledge has grown as a proportion of the curriculum. Probably that is in part because the size and the scope of knowledge demanded in the sector has expanded enormously, and in order to teach it all everything which does not seem directly useful or relevant has to be ditched, (Snyder noticed that at MIT in the late 'fifties). But it may be producing graduates (because it is certainly the case in first degrees and even on Master's programmes) who are limited--certainly at the point of graduation, although they will pick up a lot through participation in their communities of practice later.
A side issue is that the detailed specification of course objectives etc. has lead to a retreat from engagement with some of the difficult threshold concepts which are necessary for full engagement beyond the recipe level.
To a considerable extent this approach has penetrated even into the kind of teacher education programmes I have been involved with. Its instrumentality may even be responsible for the perpetuation of some of the egregious myths and fads which have characterised teaching in recent years (see Goldacre, 2008, ch.2 and here). And reading work produced for the course by some of the students who work in mainstream FE colleges is sometimes downright dispiriting, because of its lack of critical discussion--although that of others is profoundly encouraging, as they go beyond recipe knowledge.
There's a glimmer of hope, though. Proposed revision of the curriculum for ICT and Computer Science in schools has effectively recognised that ICT as presently taught is merely recipe knowledge (i.e. use of standard applications with no understanding of how they work) and needs to be complemented by learning about computer algorithms and how to code applications.
Berger P and Luckmann T (1967) The Social Construction of Reality London; Allen Lane
Goldacre B (2008) Bad Science London; Fourth Estate
Snyder B R (1971) The Hidden Curriculum New York; Alfred A Knopf.
(see also) Wolke R (2008) What Einstein Told His Cook: Kitchen Science Explained London; W W Norton