"White male, age 17–25" is an entire trait. "Homosexual" is a *part* of a trait: "Alcoholism, childhood hyperactivity, homosexuality, depression, borderline personality disorder, and suicidal tendencies". It's significant because of the other descriptors it's grouped with in that trait. It would be weird to have "homosexual" as a trait on its own, but it's obviously weird to group it with these other descriptors in a single trait.
This report [1] is specific to "Firefighter Arson", so they're looking for things to screen for during hiring for firefighters. The section you'll be interested in is on page 20: "PREVENTION OF FIREFIGHTER ARSON". They list some actions departments have taken: background checks for arson; requiring an affidavit the applicant hasn't conducted arson; "scared straight" lectures on the consequences of being jailed; and screening firefighters with a questionnaire.
But yeah, the profiling section some sound like a way to pathologize arson. I found the motivation section and case studies more interesting. The only corroboration from the case studies is for the age; there's no evidence included for the personality traits.
I count a total of 14 traits that they looked at[1]. Why is that one, and only that one, significant in your opinion?
[1] Including the trait of being white. And male.