It has occurred to me that while Wolf is built to parallel regular algorithm books, Ram is not. Ram is a collection of ingredients; Wolf is a collection of recipes. Nothing wrong with either; ingredients often require a certain preparation before use, and improperly prepared ingredients result in botched recipes. It's important to understand the components as well as the composing.
In case that seems too simple, consider stock. Quite a bit can go into stock; yet it is still an ingredient. There's crossover between the two categories, in the sense that if you're writing a filter you may well just need one or two particularly beefy ingredients ( a couple of regexes), but those same ingredients are in some other, much larger recipe where they are just as necessary, just as useful, and provide the same functionality -- in concert.
The obverse is true as well: That some simple "ingredients" like stock can be used almost all by themselves as soups. Sometimes the algorithm is the whole ball of wax; you just need a search function.
Of course, these are observations from someone who hasn't put a whole project together yet, either. I'm sure I am missing important points here.
All I'm saying is that I liked the Wolf book a lot on my first read, and will be chomping through some bits more in my spare time. I like it enough to keep rereading for a while.
Thanks, Jarkko, Jon, and John.