The first bit of the article sure makes it sound like this DNA is what does it all.
They then tried to "more realistically" replicate the scenario. Instead of just snipping out exon6, they tried inserting in various elements into the mouse TBXT gene to emulate their model of how AluY is cutting out exon6. They attempted a few different methods which resulted in different models that would end up with different relative amounts of TBXT missing exon6 being produced. I'm having a hard time tracking all their conditions and lines at this point, but they seem to claim here that higher amounts of exon6 drop out correspond to shorter tails.
They also found that they could not get any viable mice with exon6 snipped from both copies of TBXT.
TBXT itself is a transcription factor, meaning that it controls the expression of other genes. There's probably a set of genes elsewhere that together make up the "tail-making machinery" that TBXT heavily interacts with. In this sense, TBXT 'might' be the most important single point lever, but at the same time not be fully in control of the underlying machine.
What does seem persuasive is that idea that this mutation kicked off some rate of tail lose in our ancestors, which then lead to some sort of selection advantage. The advantage was then made more durable by accumulating additional mutations to lock in the phenotype. In addition, once the "tail making machinery" is inactive, we have no idea how those genes might have been repurposed. I think it's in this sense that the scientists say that we probably can't bring tails back just by "fixing TBXT".
Wait - I finally found a use case for Neuralink that I can cheer for!
Even if the driving mutation identified in the study could be undone, “it still wouldn’t bring back the tail.”
Oh well.