I am not sure what you call granularity, but the jumps are the same in 64 pitch as well. Your comment is then misleading because changing pinions of course is bound to induce large differences especially at low tooth count values, simply because of the mathematics. For instance a one tooth difference to a 20 tooth pinion is 1 in 20. The larger the pinion, the smaller the difference (for a 40 tooth pinion, the difference would be 1 in 40, of course). The result is then that if you want smaller variations (a smaller increase in ratio) you have to change spurs, not pinions. For the example above, 1 tooth in 90 is a very small difference and the result is you can hit 8.0167 (21/91) or 7.84 (21/89). You can get even finer increases if you look at combinations (spur/pinion) as a whole. For instance 96/21 will give you 8.07 FDR. And you could stick with a 21 pinion and have available 95/21 which will give 7.9886, and 94/21 which will give 7.9045 and so on.

The only difference is that 64 pitch will allow more combinations to be used before you run out of room for adjustment on the motor mount simply because of the smaller diameter increase between two spurs/pinions of neighbouring size (one tooth difference). I haven't checked these sizes because I do not run mod, so you'll have to do the checking to see what fits. At our track however, people run mod with 48 pitch without problems and their FDRs hover around the same values.