Let's simplify this the maximum:

• In the new system you have two parameters:
  a) You select the % VP you vote with.
  b) You select which value you envision the post to end up at.

Your vote can then influence the value of the post up to the max % VP you applied to it.

The closer the end value of the post is to your value, the less % of your vote is necessary, but that is not important.

Now, a user might be disappointed that he did a 100% VP vote but 0% had to be used because the post ended up exactly at the value at which the user guessed, now you could gamify this and reward the users the most curation not only who early voted, but how well someone guessed.

We don't have to use 100% of the VP of all the combined voters and we don't have to reward it back neither.