Here is how it has worked out:
Since my upcoming talk has to do with different mechanisms of identity formation, I decided to experiment with my identity as an academic and try being a stress-free, well-rested scholar who does things well in advance. I have to confess that I’m kind of enjoying it. The quality of the talk will definitely be better as a result of the writing process not being conducted in high-stress circumstances. And the funniest thing is that it will actually take less time to write the talk than it normally does when I struggle to make the deadline. When you are in a rush, papers keep getting misplaced, books keep closing the second you find the page you need, fingers that shake in panic keep hitting the wrong keys, and all ideas you might have had are getting misplaced by the hysterical, “Oh my God, I can’t believe it’s that late already.” Believe it or not, writing for two scheduled, leisurely hours in the morning produces more actual usable text than a frantic all-nighter.
I'm very happy about this, because I know the SMT approach works for me, but also need to show it works for other people in the same way.