Analytics does work and has worked for many teams for many years. The Phillies seem to have not gotten it right. I recall the first guy they brought in was an intern from the commissioner's office whose only experience was on evaluating arbitration salaries. I don't know what is wrong with either the guys in the analytics department, or what they are tasked to do, or how their work product is utilized, but something has always seemed very off. It would be a shame, and beyond that it would be disastrous to the Phillies future if they simply abandoned analytics. Like anything else, you need to learn how to get it right. We seem to have been an organization of one inexperienced true believer (klentak), an owner pushing the approach, and a whole organization filled with skeptics. This argues for the new GM to be a guy with a proven ability of success with analytics.
It would be very, very wrong to conclude that the Phillies problems were caused by misuse of analytics. We've not done any better with human scouting, and there is next to nothing that analytics can do to allow you to sign the best teenagers (and what little it can do, like show the importance of one-year age differences in rating the talent -- we seem not to do), yet we absolutely stink at this totally human-based scouting evaluation of American and foreign teenaged position players -- we misjudge both talent and character (as with the not highly motivated to excel at baseball LGJ, who blimped up within a year of our signing him). We've done best drafting college kids, where analytics does play an important role, so one could argue, with next to zero insight into the Phillies' talent evaluation process, that the analytical side of identifying amateur talent might actually be better than our human scouting side. I see no more evidence that our veteran, human, baseball lifer talent scouts are doing an even average job of amateur talent evaluation. This suggests that the 'old way' has also stopped working for us.
Since both our old ways and new way have come up dry for the past decade, it is incumbent that we hire a GM who knows how to build both a modern analytics team and a traditional human scouting and player development team and knows how to meld their input to leverage their impact on a successful outcome. The only way we can know a candidate knows is if s/he has done it and we recognize the success -- not someone who was present when a more talented GM did it -- someone who demonstrably had done it themselves.
A wunderkind like Klentak is like a high school first rounder -- lots of perceived talent, but not talent that has been verified or honed to the point of major league usefullness. The majority of wunderkinds do not become stars. It's worse with management, where pragmatism and knowing the limitations of your methodology are so important. It is someone like AOC, who is too inexperienced to know that you can't rebuild a vast economy to not use fossil fuels in only 10 years, versus an older, more pragmatic politician who recognizes that limits do exist.