OK, before this gets mainstream:
1) usability testing IS user research.
2) a 4–6 week “user research package” is WAY off, we sell 1 week interview packages (2 with recruitment), to quickly ask 3–5 potential users, sometimes over phone.
3) the reason to do upfront contextual interviews (I guess that’s what you call user research) is not to build nice personas: it’s to get your gears into the mindset. It’s to make your unconscious brain absorb all the little details about the environment, the context, the usage habit etc
4) Tacit knowledge is, and always was much more important in user research (including usability tests) than explicit findings. Personas (in my opinion) are mainly there to make it tangible for the department which pays for user research, as some of them need get “something” for their money.
5) the issue with starting design-first is that you don’t even try to understand the context beforehand. This is a very short-path to the “prove the null hypothesis” effect, which we sometimes see with Google Sprint advocates: they come up with something in a meeting room, build it up quickly the most realistically possible (usually final visuals), and go out to prove themselves (“it’s great isn’t it?” “yeah, sure”). Dangerous path.
Surely, we build difficult systems (hospital management, lawyer office systems, other non-consumer UIs), but we also have a different take on things which you’d thought obvious, which brings our clients a differentiator.
Sometimes we come back empty-handed, as is, we couldn’t find a problem which really does need a new solution, but when we do, it’s not just some guy who made something up in a meeting room (as in the case of Google Sprints) for which users said “yeah, sure”, but instead, a strong and valid hole in the world which we are able to find the perfect plug to fill it in.