There is virtually no evidence finding TBS's to be effective because there are virtually no studies that include a control group.
But there is tons of evidence that nearly every individual technique used by the "tough love" ones is ineffective and/or harmful.
For example, the boot camp model-- Justice Department has a nice 1998 meta-analysis showing it to be no more effective than prison. Same study finds no evidence to support use of wilderness.
And don't go citing the later OBHIC study-- no control group, doesn't count. there is a very nice review on their site or one that links to their research showing that the better the methodology of a wilderness program study, the less likely it is to show effectiveness. Not a good sign!
Let's look at the "personal growth" seminars-- the research on this is that they get high grades from participants for satisfaction afterwards. however, if you use a control group, you see no actual psychological or behavioral change. and, there are numerous case reports of psychological damage-- some of which have involved completed suicides and some of which lead to massive legal judgments.
Yup, those are anecdotes-- but they appear regularly enough that it would be unethical to do the controlled trials needed to absolutely prove causality since there is no evidence of positive benefit. The promoters of these seminars for regular people, in fact, now try to dissuade people on meds or with any psychiatric history from participating voluntarily which is a tacit acknowledgment of the danger-- and yet they are mandated by these programs for both troubled kids and their parents.
Now, let's look at the core of therapy even in some of the "best" programs-- the confrontational group in which teens attack each other's flaws in an attempt to "break" denial and push positive change. Guess what-- huge literature on how this backfires in the addictions field. See William Miller, see Phoenix House's own most recent manuals which discourage it based on their own internal findings, see the original Synanon study by Liebermann et al. which found Synanon was linked with the highest number of bad outcomes. 9.1% had lasting psychological damage.
Oh, and if you're still not convinced that there's more evidence against TBS's than in favor of them, look at the whole inpatient v. outpatient debate in the addictions (solidly in favor of outpatient for all but the most extreme cases), the literature on how aggregrating troubled kids tends to produce bad outcomes (see Dishion, When Interventions Harm for review-- more recently, see Szcaponik and Liddle) and the NIH consensus statement on what works best to fight teen violence and delinquency. It's community based care like various evidence-based family therapies, not inpt that has research support.
And btw, even if TBS's were effective than ordinary schools at suicide prevention, you would expect them to have a higher than normal suicide rate afterwards because you would HOPE that they were seeing a population that was far more troubled than the general school population, and that being the case, unless the intervention was more powerful than any known intervention in psychiatry, the suicide rate in that group without any treatment effect, positive or negative, would be higher than the general high school rate and only the most massively successful result would be able to return it to the level of average risk.