Another link for your collection:
How to get the most out of your mildot reticle
This gadget is worth it's weight in calculator batteries:
IMHO, the basics of mildot ranging are pretty easy to grasp, and once you've read one or two decent articles on it, you've pretty much seen what you need to see. It gets redundant, and there's really no miraculous bit of info that's gonna make it foolproof. It's like reading a ruler . . .
After that, you really need to practice, practice, practice . . . Ranging with your mildot reticle, then checking it with a laser or hi-resolution topo map is probably the best way to get going on it.
Prior to getting frustrated, you should check your reticle to make SURE the mil is a mil, and this can be done via a target marked with true mils at a known range. Good quality scopes are usually dead nuts, but lesser scopes can be wrong.
IMO, the single biggest thing is how closely you can resolve fractions of a mil with your particular scope. If you cannot get within a tenth of a mil or so, you can't estimate range much better. I have scopes with regular mildots, Gen-II mildots, a Leupold TMR and a couple NightForce MLR reticles, and I can get VERY close with the TMR and MLR reticles, next closest with the Gen-II mildot, and the regular mildots are pretty coarse for me, like trying to determine 1/8 of an inch with a ruler that is graduated in half-inch tics. . .
Another, similar method would involve using one of the many MOA based reticles, and measurement resolution is usually better. If you're starting from scratch, this might be a better way. One big attraction is that many/most scopes are adjusted in MOA, so there are no mental conversions to perform when correcting for a marked miss. If you're already running mildots in a bunch of a scopes, tossing an MOA scope into the mix often causes problems when in a hurry or distracted.
Next up would be having a target with a known size, the more exactly you know the size, the more exactly it can be ranged. If you can only know your target size within 10%, that's all the closer you'll ever get to ranging it accurately . . . knowing exact target size is crucial to exact ranging. While one doesn't always have to be all that exact, especially at moderate ranges, I find that striving to be "exact" is necessary for me to achieve "adequate" results, LOL.
Then, it's methodology. . . You need to be about as steady to mil a target accurately as you would be to shoot it, so a good position/rest is a must. If possible, mil height AND width, to help negate the effects of varying aspect angles. Align you reticle to the edge of the target, or, if irregularly shaped, try to ensure your measurement is perpendicular to the edges to be measures. Dark targets look smaller, light targets look larger, and practice will allow you to get a feel for how you need to measure the many variations. I shoot tactical matches that use steel plates at unknown distances and I will mil white plates on the inside edges, dark plates to the outside edges.
I also use mine for holdovers, and typically set up a small sketch to a given zero range. IOW, I'll set my scope to 400 yards or some other zero at mid-range to my expected application, then calculate how high and low I will be at other range intervals, and mark my drawing with POIs at those ranges. I use 100 yards increments for my flatter shooting guns. One can also zero closer and show all other ranges as holdovers, where all of your POIs will be marked below the main crosshair intersection. These methods can be pretty fast.