The Currie point method is so prevalent because there are so many folks who use forges and other non-controller run heat sources, not that I am criticizing this but instead simply pointing out that if one has a controller ran heat source one simply puts in the desired temperature numbers for the alloy; so for folks with controllers it is a moot point. If any folks get a lot of e-mails and phone calls regarding heat treatment issues, they are welcome to trade places with me anytime:3:. And I know the reason for this. We can see, with our own eyes, if we got everything else the way we want it with a knife, but not so with the heat treatment. It is this mysterious, doubt creating, nature of one of the most critical aspects of knifemaking that makes us desperate for answers, reassurance, and any kind of indicator, guideline or reference point and the more foolproof we can believe it is, the less anxiety we have. You should see me when I am working with a new alloy that I cannot get Rockwell numbers off from, I am like Sheldon Cooper trapped in a poorly maintained porta-potty.
If you follow the Fe/Fe3C phase diagram the Currie point all makes sense and works, but that diagram is based on equilibrium conditions and cannot apply to a dynamic system that involves heating or cooling without holds. It also only applies to basic iron/carbon systems, any alloying present completely invalidates it. I have done tests where I actually quenched steel from exactly 1414F, something you can't do without very precise controls, and the lower Rockwell and mixed phases were not pretty, but it would skate a file just fine.
But here is what the guys who work with forges have going for them- guys who forge prefer simpler alloys, and so they are much closer to the basic iron/carbon system described in the Fe/Fe3C phase diagram. The second thing they have going for them is the lack of control one has with such a heat source. Yes, I did say the lack of control is a benefit there:3:. With a heat source such as a forge it is nearly impossible to nail the Currie point dead on and you will almost invariably overshoot it by the time you get the entire blade up to heat. So in this case the magnet is an excellent tool to keep from overshooting the temp since you will often overheat about 50F -75F. So 1414F plus about a 50F-75F margin of error puts you right in the sweet spot for most basic carbon steels, but if you had shot for 1475F to begin with your margin of error would have you much too hot by the time you had things worked out.
Guys who use the magnet with something like 1084 heat it, check it, quench it and are happy. The forge fire gave them a pretty nice knife with that simple steel with one simple operation; with the exception of some Mn, it is that simple iron/carbon system. Now add just a little chrome and watch what happens to the simple magnet approach, suddenly folks using the magnet are finding better results by having to have a second or third go at the hardening operation. The added alloying pulls all the numbers much farther away from the system where the magnet and no soak times work so well. Now, without a controlled soak, you have difficulty obtaining full solution without taking the temperature beyond the safety zone surrounding the Currie point; rather than risking overheating, the safer route is to repeat the operation inside the safety zone until you have proper solution.
The more accurate method is to watch for decalescence, but that is something that takes time and patience to develop the skills to do, the magnet can be done your first time heating steel. In training your eye to recognize proper temperature you will then see the actual transformation as it occurs and you will develop a much closer working relationship with your steel of choice, but we are all human and you will have your bad days. The magnet gives more of a feeling of a safety net. And the controller, of coarse, is the ultimate safety net by eliminating human error altogether; it is for this reason that I encourage all makers to learn to heat treat by eye before moving onto controllers. Then when they combine their knowledge of color with what the exact temperature really looks like, they are practically human pyrometers. I have had some say they doubt I know how to heat treat without my fancy gadgets; they obviously are not aware of the years I did it by eye with a forge, and are quite surprised when I show up to a demo somewhere across the country and move and oven or a gas forge out of the way so I can demonstrate heat treating in a coal forge.
I don't have any problem with folks recommending or using the magnet since, for the reasons I mentioned, it can work. I do have a problem with calling it a perfect, or foolproof, method for determining temperature, because that just is not true. One more huge issue can be summed in one word- "hysteresis". Rather than typing another dissertation I will allow those interested to look it up.