I'll try and answer that.

PB value is position sensitive.

A great CF is worth 30-40% more than a great 1B, thus even though Pujols has around a 1250 OPS, Beltran and his 1095 OPS is more valuable.

But wait, you say, Pujols has many more atbats. This is normally correct, as I'll explain below.

One of the problems with PB values is that is has always short changed partial season guys. You could have two 1B with 900 OPS and 300 atbats and each would be given a 2.5 PB value for a total of 5.0 PB value. Another team might have a 1B with the same 900 OPS and 600 atbats and he will have a 6.5 PB value.

If teams were limited to 25 guys, then this might make sense since the number of players you can are limited thus having one guy produce what two guys produce being more valuable is correct. With 35 man rosters this doesn't really apply. Its one of the reasons you see teams with 80 PB values winning divisions over teams with 100 PB values.

I know, all the above should justify Pujols having a higher PB, or at least the same. What happens is PB values become more accurate as we get farther into the season. Currently every player is a short season guy, thus for the partial season the program must convert a players stats into a full season, thus leading to some discrepancies. This last paragraph is much conjecture on my part.