A lot of my work starts as various levels of screwing around...sitting down at a keyboard synth and poking around with various parameters until a sound pops out and says "OK...here I am, you bastard...USE ME!" At that point, I'll start looking around the studio and sorting out where complementary sound spectra can be cooked up...both reinforcing and conflicting. Eventually, there's a tipping point of sorts that appears, where the "uphill" of the messing around turns into the "downhill" of a new work coming together.
Where it gets dicey again is when I start to mix down and construct the mix's processing, gain structure, and so on....so there's another shorter "uphill" that happens as I transition the process from the synths to the desk and processors, but just like before, there's yet another tipping point and the mix just starts gelling together.
It's worth noting that this is a method I've used for decades, so I'm very accustomed to the workflow it requires as well as the nuances of the process itself. But I shouldn't think that taking the time to get used to one's own workflow patterns should be considered "time wasted". Instead, it needs to be viewed as just another method of musical practice; mixers and processors don't have "performance practice" in of themselves as a rule, so it's up to YOU to sort out that aspect of your work and make it part of your musical "fingerprint". And if it takes a week of hammering away to suss out four measures in a Schubert lied, it shouldn't be all that surprising that electronic music composition/performance/production ALSO has similar requirements.