I would be surprised if the umpire tendencies would have a huge effect on the amount of offense (or lack of offense) generated in a game. But I have had my umpires set on the "personalized" setting 99% of the time in MoM games, in CPU vs CPU games as well as in RTTS games. I did try the "perfect" umpire tendency at least once not long ago, but that was when I got called out on strikes in a RTTS plate appearance on a pitch I felt was clearly out of the strike zone. But it did not take long for me to migrate back to the "personalized" umpire setting because that is how the real world umpires call their games.
I think the challenge we all face while trying to create a reasonably accurate simmed game of baseball in MoM games is to get the pitchers to throw a combination of pitches that are both purposely in the strike zone as well a pitches to deceive the batter that are purposely out of the strike zone but at some point looked like they would be strikes and get this to happen that statistically matches the real world. Of course it is easier said than done.
I know you have said in the past that you feel the game should play well out of the box using the default "everything set to a 5" slider setting. But sometimes we need to tweak a slider setting up or down and it takes many games to determine whether or not the slider tweak was the right change to make. I am getting back into MoM games after a two year break and I am reasonably happy with the results I am seeing with my sliders that are not too far off of the default "5" values. But who knows what MLBTS20 will play like. My current MLBTS19 sliders might be trash in the new game. Only time will tell. We need nomo to work his statistical analysis magic, but has he gone AWOL this year?