It really depends on the cookbook and the publisher.
My experience is that books by people who primarily write cookbooks for a living have been tested, not always by anyone other than the author. Julia Child's first book,
Mastering the Art, was tested incredibly rigorously by herself and her co-authors. Kennedy, Hazan, Tropp -- these people work(ed) very hard to ensure that their cookbooks, written with a home cook in mind from the start, would be approachable and accurate.
When you get into cookbooks where the main listed author is a restaurant chef, you have to be more cautious. Many restaurant chefs simply do not cook at home often enough to have a clear sense of what is and is not the same. For example, most professional kitchen ovens are convection these days, whereas most home ovens are not. This means that roasting something, for example, may be listed to take 20 minutes, but if you did it at home your meat would be very underdone -- you'd need to roast it more like 30 minutes without convection.
Unfortunately, the demand for restaurant chef cookbooks is very high, but production costs are enormous. All those color glossies cost a lot of bucks, and so do the various staff (including the chef), not to mention advertising. If you want a restaurant chef cookbook that has lots of pretty pictures (much of the point, often) and is of a fair size, you have to economize somewhere -- and all too often the economies undermine the recipes themselves.
You will often (though not often enough) see restaurant chef books with a listed co-author who is a cookbook writer. Vongerichten wrote
Jean-Georges with Mark Bittman. Lagasse wrote
New New Orleans, his first cookbook, with Jesse Tirsch, if memory serves. Here it's the cookbook writer who is primarily responsible for the prose and the home-cook testing. Both these books get high marks on these points, in my opinion.
Jean-Georges reads like Bittman hung out with Vongerichten and tried to pick things up, but it's very clear from the actual dishes and results that Vongerichten wasn't just giving occasional advice.
New New Orleans reads like Lagasse did everything, but this is before he was a big TV guy and I think he didn't have all that much sense of home cooking -- yet the recipes work very smoothly. Another interesting example is
The French Laundry, which Thomas Keller wrote with Michael Ruhlman. It's quite clear that both of them put a lot into the book, and they're both extremely good at what they do -- Keller is one of the best chefs in the world, and Ruhlman is one of the best food writers. The result is an extraordinary cookbook -- if deliberately intimidating.
But you have to remember that restaurant cookbooks have other functions beyond recipes:
- They advertise the restaurant, and serve as convenient souvenirs of a memorable meal eaten at a fancy place.
- They inspire other chefs, professionals I mean: you're cooking in the boonies, let's say, but you buy a whole lot of hotshot cookbooks by the latest hip guys in New York and LA and wherever, and you get inspired by dishes, platings, combinations of flavors and textures, techniques, and so on. What you do with this may well be radically different from what the cookbook suggests -- you don't want to just produce second-rate copies of somebody else's dishes, after all -- but the book has done something useful for you.
- In some cases, they may teach cooking techniques, or present a regional or cultural style; Portale tries to teaching cooking, and Bayless tries to teach Mexican cuisine.
To top it all off, what you can do with a cookbook recipe, what I can do with it, what some of the scary folks around here can do, and so on are all different. If I use a baking cookbook, I hope it's been well tested, because I wouldn't know: I don't bake almost ever. If I am doing a Charlie Trotter or Nobu recipe, on the whole I can usually see if there's a problem, and I can expand the shorthand: if an instruction is pretty terse I still know what it means, more or less. "Make a vinaigrette from the listed ingredients" is fine -- I can do that easy. Similarly, if a cookbook tells me to spend half an hour messing around with pans of simmering water to make a Hollandaise-type sauce, I'll ignore this and do a
sabayon in a skillet in 2 minutes, the point being that I don't need the instructions to be precisely accurate because I know what I'm doing technically.
Your mileage may differ: you may need much more step-by-step with main courses, let's say, but be an ace with desserts of all kinds. Maybe you've not made much stock, and don't know how to ensure that it's crystal clear and slightly gelatinous, which in some dishes might well make a real difference. In that case, you're going to want a cookbook that actually teaches you how to make good stock, not something that gives semi-accurate shorthand.
I think the greatest problem is that there are so few sites anywhere that produce detailed cookbook reviews by skilled home cooks (or professionals who can accurately pass for this). ChefTalk has reviews, but the length is in my opinion too short: a reviewer cannot really get into the various complex layers of quality that come into play, for a wide range of readership. KY is dead-on to note that reviewers are required to test at least two recipes, something omitted in a disconcerting number of reviews, without which you really have no idea whether anyone but a professional or extremely skilled home cook can reproduce the dishes.
I'll conclude this yammering by mentioning a TV series Jacques Pepin did with his daughter Claudine. If you watch this, you really see how cookbooks can go wrong. Pepin is both very skilled and an extremely good teacher; his daughter is charming but rather unskilled in the kitchen (her husband is a chef, though, so I imagine she eats well regardless). Watching this, you can see why a restaurant chef is often a poor cookbook author: Claudine will do what she thought he told her to do, but she's sufficiently clueless and slow that the results can get in the way. For both Pepins, that's a big part of the point of this show: they want to demonstrate that you don't have to have excellent technique to knock out very good dishes. But again, he's a very good and experienced teacher, and that's not true of most restaurant chefs, many of whom don't have the greatest technique either. This is I think why so many people complain that this or that cookbook requires that you have an army of assistants. At base, the fact is that a good restaurant chef's primary skill above all else is the ability to do a great many things rapidly and accurately at the same time. What Alfred Portale or Mario Batali can do by himself in 10 minutes may take the ordinary home cook an hour and a lot of frustration -- and the results won't be as good because the home cook is desperately trying to get it done at all, where the chef has sufficient leisure in his 10 minutes to do some tasting and correcting as needed.
Ultimately, there's no magic bullet, no (you should pardon the expression) recipe for success. Your best bet is to ask around on line, to buy things on a whim and try them, and so on.