Re: Dear Matias,
@readinthereg - I think you may be missing the point I picked up on, hence why the down votes.
Going right back to your original statement: "there are some genuine questions as to whether we can create a better interface between us and machine". I doubt anyone is actually questioning the validity of that statement, only how we go about answering those questions, in the context of an established major product like Windows, Android, OSX, iOS, Linux, z/OS etc.
So yes, let the designers play - just as we let pharmaceutical research scientists play. However, a big difference between pharmaceutical research scientists and software developers is that once the pharmaceutical research scientists identify an interesting compound, it then goes through an extensive, regulatory evaluation programme (well unless it is a possible treatment for Ebola) before it gets anywhere near the market. The release of Windows 8 (based on the stories of the key engineers) was a bit like a pharmaceutical company withdrawing an established product from the market and replacing it with a product that shares a common parentage, but hasn't gone through the full evaluation programme and in fact has gained negative results from what limited trials have been performed, just on the word of a research scientist. Win 10 is simply the pharmaceutical company trying to alter the new compound because it doesn't want to admit a mistake and resurrect the old product...
As for you point about focus groups, that is down to knowing your product and its potential customers. Apple with the iPad conducted extensive research using very a specific focus group, namely young children, additionally, they did much to include really good accessibility features. However, in later versions of iOS many of these features (and strengths) were diluted as design school purist idea's took over. Also, they did make some glaring mistakes, such as the weird 3-D floating icon effect, but you could simply turn if off and forget all about it.
Finally, we can forgive the DH engineers because they were working on the edge of the known and hence had no prior indication that their windows design was fatally flawed. However, there is plenty of evidence to show that UI/UX workers (specifically at Microsoft) weren't and in fact ignored feedback that was telling them their ideas were daft.