I mix different Lot#s of powder all the time. It works just fine. The key is to be sure the powders are really mixed well. I use a 5 gallon white paint bucket I bought at Lowes with a snap-on lid that has a rubber O-ring. I layer the powder into the bucket, one pound at a time from each successive Lot # that is going into the mix. So the "mixing" process actually starts as the powder is added into the bucket. Once all the powder has been added, I tape the lid down securely (i.e. packing tape going all the way around the bucket). Then I shake the bucket back and forth (left and right) about ten times, then roll it about ten times. I repeat this process multiple times during a day, usually whenever I walk by the bucket sitting on the floor. At the end of the day, I call the "mixing" process good, and distribute the powder mix into empty 8 lb jugs.
I would suggest recording the velocity for the current load as a starting point for load development with a new Lot# of powder (or mix). In my hands, reproducing the original velocity will usually get me close. Nonetheless, this approach by itself does not allow one to know where they are in relation to an optimal window. In other words, you may end up inside the optimal window simpl;y by reproducing the original velocity, but you really don't know where the load is within the window. Is it right next to one edge, just waiting to go out of tune when the conditions change? Or is it smack in the middle where you'd like it to be? Without further testing, there is no way to know. Likewise, re-doing a seating depth test is not that big of a deal. It is unlikely that one would have to cover the same wide seating depth test range they may have covered when the rifle was brand new, as the seating depth optimum shouldn't change a great deal with a new Lot# of powder and a load adjusted to similar velocity using it. But it might change a bit, so testing two or three .003" increments to either side of the old seating depth is really not that painful and then you know with certainty. So when you have an established load, re-producing the original velocity may be a large part of re-producing that load with a new Lot# of powder, but a modest amount of further testing will really allow you to optimize it without a lot of guessing.