With modular (as with most electronic music tools) I find myself focussing on timbre much to early in the composition process. In an age of one-minute Instagram videos, I find myself chasing novel sounds that I hope will catch people’s attention and inspire my next great work. When I arrive at a timbre I am pleased with, I become so satisfied with my achievement that I neglect melody, harmony, rhythm, dynamics and form. After some time has past, I am left feeling empty because I conveyed an idea, but never told a compelling story. It has taken me much too long to realize this, and I am finally adjusting how I write music accordingly.
An old trick of great orchestral composers is to suss-out the melodies, harmonies, rhythms, dynamics and form on the piano first, in an attempt to focus and as an alternative to endlessly exploring the timbral capabilities of the orchestra before a solid musical foundation a laid.
A more modern trick is to sketch songs/pieces on a shitty mono speaker because if the structure is compelling enough, it should sound good on any speaker, and the fancy beautiful timbres that are added/subbed-in/brought-out later in the mix will just be icing on the cake for the more high-fi listener.
Find a process that works for you and understand why certain limitations can be a gift.
I would rather attempt to write the most interesting piece of music on least-compelling piece of gear and fail than succeed at making the least-compelling piece of music on the most interesting piece of gear.