Stockholm syndrome has been discussed in relation to domestic abuse, but not all that much. In the Encyclopedia of Domestic Violence (Routledge, 2007, NA Jackson ed.) I found criteria given that mirror those for kidnapping victims:
- The victim perceives a person threatening her
survival. The threats may be physical or psychological.
It is not important whether others
view her survival as threatened, but rather
whether she does.
- The victim perceives the abuser showing her
some kindness, however small. For example,
the kindness may be that for one day out of
the month he does not abuse her.
- The victim is isolated from outsiders. This isolation
may be physical—she is not permitted to
have contact with family or friends—and/or
ideological—she is permitted exposure to
only the abuser’s perspective.
- The victim does not perceive a way to escape
the abuser. Batterers use violence to help ensure
that their partners do not leave them.
These are pretty strong criteria to fulfill, particularly number 3 and 4. The Encyclopedia doesn't have any systematic evidence how often all these are fulfilled, in other words we don't know what the prevalence in the population would be. There has been one scale proposed by Graham et al. (1995) for measuring a more gradual version of the syndrome, but I haven't found other papers using it.
As for the usual reasons why the victims remains in an abusive relationship, they usually see (or hope for) some redeeming quality in their abuser:
a considerable number of women felt their abusive male partners still possessed some good qualities: More than half (54 percent) saw their partners as highly dependable, while one in five (21 percent) felt the men in their lives possessed significant positive traits (i.e., being affectionate).
What you are proposing is a bit different than Stockholm syndrome. When you say that
a small portion of the population that struggles to open up to and connect with potential partners tends to form bonds with abusers
you're basically proposing is that some victims have a propensity (perhaps a personality trait) for abusive partners. I'm not sure if that's the case; surely some women have submissive or masochistic BDSM fantasies, even rape fantasies, but to argue that they could only bond with abusive partners seems extreme to me, even a form of blaming the victim. I haven't seen this trait-lie preference for abusive partners discussed in literature. As cited in the Encyclopedia Graham in a 1994 book (Loving to Survive: Sexual Terror, Men’s Violence, and Women’s Psychology) did propose that victims may develop "borderline-like" personality characteristics and behaviors... but again he says develop, not have them beforehand.
There is however one recent (2008) paper that does discuss the personality of the victims:
Women victims of IPV [Intimate Partner Violence] had higher scores than controls in schizoid, avoidant, self-defeating personality scales, as well as in the three pathological personality scales (schizotypal, borderline and paranoid). Both physical and psychological IPV were strongly associated with personality disorder symptomatology, regardless of the effects of childhood abuse.
This is after the fact though. A longitudinal study would be able to establish correlates before the IPV, but I'm not aware of such a study.