Interesting question, but it is not exactly a yes or no type answer, as it relates to the accuracy of the prediction.
The general problem with using the Greenhill formula, as in this example, is that the Greenhill formula is a simplistic tool, designed to predict the twist required to stabilize a projectile of a particular type, with a specific density, within a given velocity range. (I say "projectile" because if I remember correctly, Greenhill may have been primarily concerned with artillery...)
Anyway, since plastic tipped bullets have a different density in relation to length, the formula wouldn't apply directly to them. Even using the formula on match hp bullets is a stretch, because of the airspace in the nose. Add to that the fact that muzzle velocity is a factor in bullet stability, and the formula does not have a provision to adjust for that, and you can see how unreliable any figures generated might be.
It is a logical approach, and I have considered it, but compensating by measuring only the "substantial" portion of the bullet might not work well either. The formula is not complex enough to allow for any signficant variations, and the result would really be more in the realm of a "guess" than a reliable figure.
I would look at actual real world results over relying on Greenhill's formula to determine barrel twist at this point. The numbers that the formula spits out generally do fall in the range of stability, but are often overkill...if you pardon the pun.
TC