OK, I stand corrected.
For some reasons, I thought that there were an option to describe after-the-fact singles into "bunt for singles", something similar to "improve out distribution" but related to hits, but I believe this was entirely from my imagination.
I went at length through boxscores, roughly 60 games, and all singles I can see are:
singles (lf), single (cf), single (rf) except when singles come def-x. In other words, all singles are described as if they were hit in the outfield, so clearly, SOM does not bother to take some singles and describe them as infield singles or bunt singles.
Further, no "bunt for single" have readings. So all the bunt-for-single that we see in the description come from the utilisation of the "bunt" option available to HAL.
I found one that one guy made an experiment with Ellsbury, back in those times where Ellsbury was a A-bunter and 1-17 runner:
BUNT FOR A HIT
The "Bunt for a Hit" rule, which TSN uses, allows much more successful bunts than I thought initially. That rule depends on the speed ability of the bunter, so the ideal bunter is a A-bunter 1-17 speed.
In fact, one guy I found on a website who simmed an entire season with J.Ellsbury using extensively the Bunt-for-hit rule. When forcing defense to play back, Ellsbury on-base was over .400. With defense playing corning in, which decreases Ellsbury bunting ability to B, on-base was still at .336.
Fortunately, HAL draws the corners in for all speedy A or B bunters. Still, this move means that some outs to third and first basemen become singles. The advantage is not much (.005, so that a .310 hitter becomes a .315 hitter), but you might still consider it when building your lineup.