MS is talking about simulating an ASIC almost two orders of magnitude bigger against arbitrary workloads of various lengths.
Rather than one machine, Microsoft were running their simulations on multiple, and over a long period of time. They surely were cranking away at the work. But we know that a large portion of their total testing was to ensure backwards compatibility wasn't broken, not to drive new hardware designs with unique benefits. And given what they said about their changes, the remainder doesn't seem to have concentrated on that path either.
At this point, I should pause to clarify the point I'm making. Let me be clear that I'm not here to downplay the amount of effort involved from Microsoft. They put in a shitload of work, and it shows in the results, One X being a great machine. What I'm saying is that all this work wasn't tied to a fundamental, transistor-level rearrangement of what AMD had already done. It was more the standard process for their semi-custom business: a customer choosing from among available AMD tech and balancing it as desired. (Along with some minor logic changes to better support the profiled needs.) Therefore, repeated enthusiastic language by either partner about collaboration, as now with Scarlett, isn't necessarily an indication of a different approach.
I think it a bit naive to assume they disclosed all their customizations, or even all of the significant ones.
"All" definitely not. But it seems even more naive to me to believe they'd specifically tailor their initial PR push around how much customization they did, and then only mention the least-important, least-custom changes they made.
That is a tremendous undertaking and an investment you'd make only if you intended to effect serious change to a device.
Microsoft said they spent $100m updating the Xbox controller, and the result wasn't a powerful restructuring of how we think of or interface with gamepads. It's very possible to spend oodles of time and effort on business questions, just to end up with a boring (but effective) answer. I've been involved with such projects myself, and I'm sure you have too. Indeed, I contend it's safe to assume that
every project by a major corporation has immense input of man-hours and skill. Even the ones that end up stupid or shitty. Effort doesn't equate to novelty.