There is a way to make up for poor play in the adjustment, but that is a maximum of .10. That covers elements not in the system like the quality of opponents (above the replacement level), highlight blocks, and high value special plays (like Zuttah's fumble recovery). Some people were very vocal when I began to add this adjustment, because they didn't think it should matter whether the opponent was Albert McClellan or JJ Watt, you still have to block him. In the charts, I show the adjustment in red and raw score in purple.
I keep all of the information on pulls, pancakes and level 2 blocks. If those are highlight blocks, it contributes to the adjustment.
Pass blocking is all about minimizing failure rate, so that's by far the most important component of score.
Run blocking isn't as important as pass blocking because there is less swing in win probability on those plays (lower leverage). That is something I have discussed with PFF about including in their system, because the information necessary to calculate leverage is there. The FO addage: "Teams run because they win, they don't win because they run" has never been more true in the NFL.
There are a number of ways to consider the costs of penalties, but win probability is probably the best. Since the defense needs big plays to stop any drive (said otherwise, they can't stop any NFL offense by allowing their avg yards per play), chunk negative plays like penalties have a significant cost. I charge -3 points per 5 yards of penalties, which I have considered reducing to -2. If I made that adjustment for Zuttah, he'd have 4 more points and a D.