The burn rate of a given powder is only one input that is important for obtaining accurate and useful predictions with QuickLoad. For example, the energy content of a given powder is also important. Double-base powders will typically give significantly higher velocity with a given cartridge/bullet combination than a single-base powder of comparable burn rate because they tend to be much higher energy powders. The bottom line is that if the manufacturer is unwilling to provide the information, the best you can do is improvise.
Along that line, users of QuickLoad will be familiar with the fact that the initial input data for various powders, bullets, and cartridges all come from factory presets in pull-down menus. The key here is to remember that it is simply not feasible for Neco to provide different powder data for every possible cartridge in which it might be used. Yet factors such as cartridge volume and shape, bullet weight and design, etc., can have a very significant impact on the outputs QuickLoad provides. For example, a powder burn rate can change noticeably depending on pressure, which is heavily dependent on the specific cartridge and bullet weight loaded. Instead of trying to cover an almost unlimited number of possibilities, Neco has setup pull-down menus for each powder with some factory-preset data, then allowed the user to modify the data files so that the outputs better match their actual measured results (i.e. velocity with a given charge weight).
One potential solution for missing powder data files would be to select an existing powder file with a similar burn rate and energy profile to the powder you want to create a new file, then modify the burn rate until your predicted velocity with a given charge weight exactly matches the average measured velocity. Without having the missing powder data, you can only make an educated guess as to which one to actually choose, but a good place to start is by selecting a powder with a similar burn rate range, and one that is either single-base or double-base, as the case may be. The results of this approach will not be perfect; you're making some assumptions that may not be correct, or may only be partially correct. Then again, even the best QuickLoad files made using all the appropriate inputs are not perfect, and must often be tweaked regularly as external conditions and/or the load itself change. Nonetheless, it may be a way to obtain some useful data from QuickLoad in the absence of having the actual powder data.
It's worth noting that even where we have the actual powder file for a given powder, QuickLoad outputs are only predictions. In terms of predicting the actual results we measure, sometimes the predictions are very good, sometimes mediocre, and sometimes they can be off by a significant amount. As is usually the case, the better the inputs, the closer to reality the prediction will be. If I were using PP2000MR or PP4000MR, I would certainly give this approach a try, because you really have nothing to lose as long as you treat the prediction with proper caution/skepticism, until you have measurable data to determine how well it matches what you actually observe. The other nice thing about doing these sorts of in silico exercises is that if one particular powder chosen to start creating your own file to mimic a powder for which there is no file available doesn't seem to match up very well, you can always pick another one.