If you read on B-R.com, you'll see that similarity score only has a limited amount to do with OPS, but looks at a bunch of counting stats and a couple of percentages.  So that you don't have to go anywhere to read it, here is what it says:

Similarity scores are not my concept. Bill James introduced them nearly 15 years ago, and I lifted his methodology from his book The Politics of Glory (p. 86-106). To compare one player to another, start at 1000 points and then you subtract points based on the statistical differences of each player.
Batters

    * One point for each difference of 20 games played.
    * One point for each difference of 75 at bats.
    * One point for each difference of 10 runs scored.
    * One point for each difference of 15 hits.
    * One point for each difference of 5 doubles.
    * One point for each difference of 4 triples.
    * One point for each difference of 2 home runs.
    * One point for each difference of 10 RBI.
    * One point for each difference of 25 walks.
    * One point for each difference of 150 strikeouts.
    * One point for each difference of 20 stolen bases.
    * One point for each difference of .001 in batting average.
    * One point for each difference of .002 in slugging percentage.

To this there is a positional adjustment. Each position has a value, and you subtract the difference between the two players position. James just uses primary position, but I computed an average position for players who had more than one primary position. (See Ernie Banks)

    * 240 - Catcher
    * 168 - Shortstop
    * 132 - Second Base
    * 84 - Third Base
    * 48 - Outfield (James distinguishes, but I don't have that data incorporated at the moment)
    * 12 - First Base
    * 0 - DH

The reason that Carlos Lee is the 3rd most similar batter is because he is 4 years older than Pujols and has been playing longer.  I would not put Lee in the same class as Pujols, nor would I put Geoff Jenkins there (his number 2 most similar).  In fact, I really only look at similarity score by age if I ever look at it for someone that is still playing.  For instance, Kevin Millar is Pujols 8th most similar batter?  But oh boy does Pujols have a nice list of similar by age.

Either way, even Bill James admits that it is just a toy.  It's a fun way of looking at players and comparing them to the past, but it doesn't do anything about adjusting for eras or parks.  It's a nice way to start an argument, but it shouldn't be used to end one.

And as for your aversion to OPS.  I hope you realize that no one believes it to be the tell all stat.  The beauty of OPS is its simplicity for calculation.  It tells a lot about a player in one simple number, but obviously it doesn't tell you everything.  However, if I know that I have a player with a .950 OPS and that's all I know, well, I can take a pretty good guess that it's a player that I want on my team.

I started collecting baseball cards in 1985.  Tops sold the 15 card wax packs and a clear plastic pack of about 50 cards.  The Biancalana card I'll always remember because of the name.  But his picture shows a tall lanky white guy weakly tilting the bat backwards as if he's barely able to hold it.  Sadly, a .205 career batting average is a fitting caption to the photo.  

Biancalana was traded away from the Royals in 1987 to the Astros for Mel Stottlemeyer.  Huh?  No, no . . Mel Stottlemeyer, Jr., who was the first round pick (3rd overall!) in 1985.  He made it to the majors for 13 games in 1990.  His brother Todd was a different story.  Buddy was out of the game at the end of the '87 season at age 27.  

I also didn't list that Garth Iorg ended his career in 1987.  I had Garth's '85 card as well as his brother Dane's.  I didn't realize they weren't the same guy until years later.  

But my favorite 1985 card was Rusty Kuntz . . . 20 years later and the name still cracks me up.