I think it makes more sense to see it in the same line, not as a 180 reversal: They got PowerPC because it was a good solution at the time and the 68k they used beforehand couldn't keep up. Then 10 years later (during a time where that meant a massive increase in speeds etc: 10 years of PowerPC went from 60 Mhz to 2.3 Ghz) they saw that IBM couldn't keep up with Intel (and small unit numbers meant that it was expensive) and switched platforms to then-strongest choice. Now it's 15 years later, they already have expertise and designs for ARM from the mobile market, Intel is less clearly the leader in x86, ... and they are reconsidering again. To me, the connecting line is more that they are willing to break the ecosystem to follow wherever the best option is.
You could see it as an reversal in that they go from niche (PowerPC) to mainstream (Intel) to niche (nobody makes ARM desktop systems), but I think that ignores that they already have leading expertise with ARM, ARM is not a niche overall (tooling support and ecosystem), they're not bound to fab tech of a specific company anymore (because they now can have whoever is the best in fabrication right now make their designs, that wasn't really a thing in PowerPC times), ...