When I want to explore new sounds, riffs, etc. I put Scaler first, driving an accompaniment plugin, then I eventually duplicate and sync it for the other instruments
When I want just jam, or I want a fast and particular rhythm, for example a blues or a funk, I prefer using the Toontrack’s trio that is faster and easier, but when it comes to solos I use Scaler to have possible scales suggested as explained here
So, to me: Scaler first or Scaler after, as long as Scaler is there
My latest project is a single piece with four distinct sections, and I have Scaler driving about 1.5 of them. Lately I’ve been taking a set of performance figures and phrases across a progression, separating out the bass, mid and higher notes and assigning them to different VST instruments. And sometimes filling in or thinning some of the melodies to create space. I’m really sensitive to over-reliance on Scaler for melodic ideas, although the last section of my current piece (almost finished) has three different instances of Scaler. I’m pretty close finishing it, but in terms of clarity of workflow, it’s a dumpster fire. And FWIW, I never select a key/scale or canned progression in Scaler when I’m starting out. I’ll start with a single chord, then add, add some more, swizzle and experiment. In terms of harmony, I’m usually walking into it backwards and blindfolded.
Studio One has a chord-detect function, and dropping in one of these performance pieces from Scaler and switching it on in Studio One often renders completely different chord assignments than the ones I selected Scaler. At that point, if I want to harmonize, I have really look at the notes and try things out. My use of Scaler as a tool (still foundational) has really changed in the couple of years that I’ve been using it.