IMO it just means Microsoft is less picky about the GPU arch and they'll use whatever AMD has available at the time.
That's why Scorpio ended up with a Polaris GPU and Neo with a 2xFP16 arch closer to Vega.
At this point, Microsoft is probably stronger on the software front and Sony is more demanding on their hardware.
Microsoft could e.g. just get a Vega derivative iGPU whereas Sony demanded significantly more customizations, to the point of AMD calling it a new uarch.
In the end, I think it all boils down to the preferences of their 1st parties. Sony's ICE team could have lots of particular demands whereas Microsoft may just get more from raw power out of their hypervisor/virtual machine implementations.