The way I look at it, is that "doing things right" is a risk. It will always equal faster "speed-to-market" in the long term. I think we're just afraid to admit that "doing things right" very often ends up not working out. At all. If only it were so simple, that we could just take a little bit more time and have that perfect architecture and test coverage. In reality we often end up making something that's not "right" at all. Or worse - nothing.

I don't see taking on "technical debt" as being faster. It's just less risky. You're more likely to have something that does something after that month of work. Hence the common perception that it's faster. In an organization where there's mistrust (rational or not) it's easier (or even better) to take the safe road of "technical debt".