Who wins more games

Assuming you mean when Kershaw is healthy, that lineup would dominate baseball

Who wins more games?  

What's the game?  I'm not clear on how this game is played.

If Roster A pitches to Roster B, I'm thinking Roster A gets the upper hand.  But whether they hold Roster B to 1 run or to 10 runs is meaningless without context.  So for context I'll assume that Roster B then pitches to Roster A. In that case I say Roster A lights them up.  So Roster A is your winner.

Depends.  Does Roster B's manager let Trout steal bases?

On the hitters game of failure theme, if these great hitters  fail 65/70% of the time against standard 4/5 team rotations, what would they do against these guys? Not good I'm willing to guess

If we're playing "real" games with however many doppelgangers we need, roster B destroys roster A because Roster B has guys who've actually caught and are capable of playing above average OF defense, and passable IF defense at the MLB level. And, assuming one of them can throw a breaking ball, I suspect they outhit Roster A, too.

Absent the breaking ball, though, it would be interesting to see if Bumgarner led this "league" in HR.

Also, anyone interested in this hypothetical will probably enjoy this link, http://www.amazinavenue.com/20...ue-of-bartolo-colons.

The closest you will get to a legitimate answer is add up WAR for both lists. Anything else would be even more hypothetical.

For further hypothetical discussion

fill out the rest of the roster with your average players

for example for the hitters teams add any No 3 starters

for the pitchers teans add any batters not 3,4 or 5 in the lineup

I think the hitters do a little better considering that hitting is almost 100% of run scoring but pitching is only like 70 percent of run prevention the rest is defense. 

A starting pitcher going every 5th day actually has an ever so slightly bigger impact than a hitter playing every day buy if you consider defense the hitter comes out on top comfortably (in hs and college that is different since there are more off days the impact of an ace is clearly bigger than the one of a dominating hitter because the off days mean he pitched every 3-4 games instead of every 5th- that makes it understandable the pitchers are often prefered in scholarships over hitters. 

But in MLB if you look at WAR most of the time the leader will be a hitter. 

Had to make a couple changes, subbed Posey for Rizzo so the hitters wouldn't have 2 1B, used a collection of various randoms as the "average" players, and where possible duplicated the randoms for both teams.  So, both teams had the same 7 relievers, the same LF (Brett Gardner), the same SS (Asdrubal Cabrera), etc.  "Hitter" team rotation was Fiers, Hammel, Colon, Koehler, and Happ.  All based on 2015 numbers, run through a 162 game season with DMB.

Hitters went 101-61 the first season I simmed. Hitters best SP performance was probably Colon's 17-5, 3.02.  Glen Perkins went 2-3, 3.30 with 26 saves as the closer. Hitters best offensive performance was Harper's 310/417/632 with 48 HR, 136 R, 98 RBI.

Pitchers best SP performance was Kershaw's 8-14, 3.99. Perkins went 5-3, 3.04 with 17 saves for them. Best hitting performance was probably Adrian Gonzalez' 238/298/400 with 24 HR, 57 R, 88 RBI.

Results were much the same on additional runs. Stacking offensive studs, apparently, is more valuable to generating runs than having a stud SP is to preventing them.

Love the analytics Jac!  Very interesting. Game is certainly more complicated than we all think

Don't forget that position players don't only score but also prevent runs.