These may seem like odd questions, but I must ask. First: are they new hammers that may have been arced incorrectly? Second: What size grand piano is this problem persisting in? The length of the keystick has a lot to do with checking quality.
Rote learning is defined as the memorization of information based upon repetition. If you only have data as a tool to solve problems, you may find yourself caught in a feedback loop. All the numbers presented to you may not solve your problem if you are dealing with unusual circumstances, and I suspect your problem falls into that category, i.e., new hammers installed on a short piano.
Understanding the physics and geometry of the action is necessary to solve your problem. Watch the action of the backcheck in relation to the arcing of the hammer tail. If you find that the backcheck "checks" on the upswing while giving you proper checking on the downswing, this suggests the arc of the tail is improper (too great) for a short keystick piano. Remember, backchecks mounted on short keysticks travel at a faster rate toward the arc of the hammer and they will "bump" into each other on the way up. If you have a large arc on the hammer tail, they will again, "bump" into each other on the way up. You may find yourself trapped into a corner where it would not be so evident on a larger piano with a longer keystick with a larger hammer arc. All the adjustment up or down or bent forward or back may not solve your problem. You must carefully examine the travel of the two action parts while visualizing what a "proper" arc may result.