You have to go back to the early goals of the projects for "next gen" gaming machines of each company.
MS wanted 8GB of RAM no matter what and they were just off the buzz of the release of the original Kinect, so they picked DDR3 + ESRAM as a realistic goal (around 2010 probably) then they built around that with AMD, assuming 8GB of RAM would be important enough to give them an advantage in games. MS also added a few extra processors to aleviate the CPU and GPU limitations (the move engines).
- The problem is that the ESRAM located on the CPU/GPU die takes a lot of space that could have been used for a faster GPU (like the PS4 has) the problem is that the DDR3 memory would be too slow to feed a faster GPU anyway, so they need some sort of buffer...
Sony wanted a fast GPU with a pool of very fast memory (along the other tweaks they requested from AMD on the GPU) at first rumors hinted at 2GB, because of the high initial price of GDDR5 memory, then rumors started circulating that it would have 4GB ... and they announced the machine with 8GB last Februrary, because yields and price permitted it
Thus Sony created a machine that would release at the same time as zero hardware deficiencies as far as games are concerned, the Sony and MS engineers are probably just as good as each other, however the people who made the decisions of what they wanted in the consoles asked for different things and we got the desired results on both case.
The xbox one is more a business model than a hardware platform... it's not a consumer product, but a money printing machine for MS... that's how it feels anyway.
