http://books.google.com/books?id=zV9bpy ... e&q&f=true
I was surprised by how many enzymes have been documented in malt, the very wide temp ranges over which the various enzymes actually work, and the roles it is believed each of the enzymes play. A couple of the things I learned:
1. Alpha-amylase is considered a dextrinizing enzyme, not a saccharifying enzyme. That is, it breaks down the larger amylose and amylopectin starch chains into largely non-fermentable dextrins. Beta-amylase takes over from that point and further reduces the dextrins into simple sugars. As such, while alpha-amylase may be optimally active at a higher temp than beta-amylase, they, by definition, always have to be working at the same time for any sort saccharification to take place.
2. It's apparently possible to get a single infusion mash to convert completely by doughing-in as low as 120F, since alpha- and beta-amylase are active down there. The mash will take longer to convert than at higher temps, but it will still convert. I never would have guessed this, considering plots like that given in How To Brew of the optimal temp and pH ranges of various brewing enzymes.

3. Starches do not have to be gelatinized to be broken down into dextrins and simple sugars. Dough-balls are still no good, since the enzymes need a liquid substrate to act on the starch, but the gelatinization temperature is apparently just short of meaningless for properly hydrated malt.
4. The complex interplay between pH, temperature, grist composition, water mineral content, and mashing schedule as they relate to the final product is not completely understood. The concept of an "optimal" temperature or pH has been estimated via completely artificial experiments -- i.e. where one enzyme or another has been allowed to function in isolation. Applying these results to a typical mash is not helpful since the interplay between enzymes is not considered in the "individual optima" experiments.
5. While it would be accurate to say that brewing science is a mature field, there is still a lot of research going on to build a more detailed model of what takes place inside the mash, boil, fermentation, etc. The fact that they can't predict something as simple as the final fermentability of a beer, based on known, controllable conditions (e.g. grist composition, mash temps, boil duration, yeast pitching rates, fermentation temperatures, etc.), makes the practice of this field every bit as much an art as a science.
It could be that I just haven't read the various homebrewing texts very closely when they discuss the scientific side of brewing, but I have been repeatedly surprised by how different a picture this brewing textbook is painting than I've previously understood. For example, homebrewing literature often indicates that a "protein rest" is unnecessary and that it might do more harm than good, while this book doesn't even bother to refer to a rest around 115 - 135F as a "protein rest". It is typically taken as a matter of course that some type of a rest will be had in that range, regardless of the degree of malt modification.
For anyone that is interested in learning more about brewing science, I can't recommend this brewing textbook enough. It makes my brain hurt periodically, but it's completely worth trudging through if for no other reason than that the perspective is just so different from what we homebrewers typically see in books, magazines, or forums.