My opinion that xwOBA is the best predictive stat available is not "all of a sudden". I have literally based most of my views about players on that stat for several years now across hundreds of posts. You may want to consider sample sizes when quoting stats like wRC+, which are much more volatile than stats like xwOBA. Hint: the sample sizes for Shaw the last 2 years are not very large.
Again, I don't think anyone is championing Shaw as the strong side of any platoon. I literally wrote in the post you quoted, "The idea with Shaw was not to give him 400+ PAs vs RHP unless Riley fell flat on his face. The idea with Shaw was to add a cheap MLB hitter that could compliment Riley a bit, and potentially take over if Riley regressed. Riley would still get his shot early in 2021 to see if he continued to improve."
The choice comes down to Camargo at $1.4M as a terrible backup to Riley vs Shaw at ~$3M(?) as a competent backup to Riley (especially vs RHP). I know what I'd choose if the payroll is there, and the choice would be easy. Riley deserves a chance to continue improving, but I sure as heck don't want Camargo as the backup plan in case Riley regresses after ~200 PAs in 2021 and is dragging down a team with championship aspirations.
Additionally, Shaw is a "for example" name thrown out there as a cheap and competent compliment to Riley to serve as insurance at 3B. There are probably other names I can't think of that would fill that role as well. Camargo isn't even that, and while clv has a point that Camargo can serve as a utility IF for a poor team, the hope is that the Smyly and Morton deals indicate the Braves aren't a poor team that needs to suffer with a guy like Camargo on the roster. If the Braves can't afford a competent corner IFer as well as a competent backup SS on the bench, then AA probably messed up giving Smyly $11M so early in the off season.