While somewhat subjective, I think the "reliability" you mention isn't much higher in the 73 than in anything else. You have to watch pretty much every plane. Very few VNAVs adequately take into account the algorithm for altitude capture, and most draw that stupid present position vertical path after the first restriction unless you execute a workaround (all fleets have them) and trick it to do what you want. Strong tailwinds in the descent are a challenge for every box I've flown or observed from the JS. You always have to do the math and be prepared to intervene. ASAP report summaries very clearly show examples of automation shortcomings in every fleet.

But getting back to the original topic of why its desirable in new hire classes, I don't think very many people are basing that decisions on any kind of factual data (excluding the obvious picks for basing etc). It has nice looking glass displays, and to some (not all) that is an extremely important thing. So much so, that some would potentially lose out on a base they want more or perhaps faster relative seniority out of fear of the 88 or out of admiration for another plane that are both highly exaggerated.

We've all heard the joke about the pilot that just can't let go in retirement so he sits in a linen closet with a vacuum cleaner running right next to him. That pilot must really miss the 73.