I've been using mixed headstamps with no problems. If I'm experimenting with a new load, weighing each charge etc., then I will pick out a number of a particular headstamp for the test. After that, I go back to mixed.
I tend to separate them after loading and use the matched brass in competition. I've never seen any personal proof that this is necessary at all, but it makes me feel better. I also have a tendency to relegate Remington .45 brass for practice only since it has less neck tension than anything else I've seen (determined by the effort needed to seat the bullet).