I do have to agree with you insofar as the studies' particular results go: in both scenarios, the patient suffered: in the first scenario because of a lifetime of depression, in the second because of the same + the trauma of the early life experiences. So, if we follow your dictum:
> We feel empathy for what they went through.
(which I mostly agree with) —then yes, the relatively higher portion of empathy should go for the character in the latter scenario (which matches the results of the study).
That said, if we generalize the study and apply it to the setting it's concerned with (medical practitioners empathizing with patients)—it's not a matter of relative measures between multiple patients; it's a question of whether the empathy a patient may (or may not) inspire is sufficient to cross some threshold for a medical practitioner.
If the fact that the cause is of biological origin is sufficient to drop the empathy response beneath that threshold —then it's problematic (the paper points out supporting research for just how problematic it is). After all, the person who has suffered a lifetime of depression does deserve empathy from a doctor/therapist, even if not as much (I guess) for somebody who suffers additional afflictions on top.
Actually that makes me wonder how much dealing with many patients who are very badly off (and deserving much empathy) moves the threshold for the doctor, so that maybe they can't empathize so much for patients with situations which are legitimately bad—but not as bad as what they've seen a lot of...