It's unfortunate that Alliant doesn't list starting load velocities, instead just telling you to knock down their listed load 10% and proceed.
In the 1995 Precision Shooting Reloading Guide, Dave Milosovich developed fixed velocity .308 loads using IMR 4895 and 4064 under 180 grain bullets. Below 2400 fps it took more 4895 by weight than it did 4064, indicated 4895 was slower (more grains needed for each ft/s of MV). But at 2400 fps the charge weight was essentially identical, indicating the burn rates were the same. Above 2400 it took more 4064 than 4895, indicating it had now become the slower burning powder of the two.
So, burn rate is affected by pressure and temperature. Burn rate charts can tell you how two powders compare under the standard burn rate test conditions. The do not tell you how they compare under other conditions.
In the case of the Milosovich test, the number of grains of powder needed to achieve different velocities were straight lines for both powders, but lines that crossed at 2400 fps. If you have the velocities produced by both the starting and maximum loads that are published for a powder and a particular case and bullet, then you can figure our which is burning more slowly at those conditions. Just divide the difference in starting and maximum velocity by the difference in starting and maximum charge. This gives you ft/s/grain (aka, fps/grain) whichever powder has the lower number is having its burn rate less affected by pressure and temperature and is behaving at the slower powder. Just be aware the roles might reverse in a different chambering or with a different bullet weight.
Incidentally, the powder with the smaller calculation result, above, will also be the least affected by other factors that change temperature and pressure and thus is likely to have a wider sweet spot range when you find a tuned load for it.