I don’t work in game development professionally, so I wouldn’t claim direct industry experience there.
My comment was more from a general design/prototyping point of view build the mechanics first with neutral or placeholder graphics, then develop the visual style once the gameplay is working.
If the shell is already built, then the next stage is probably not just “what should it look like?”, but what makes it different enough to be worth playing or selling. That could be the obstacle behaviours, the level design, the scoring system, the theme, or the way the physics is used.
That said, like most people here, I don’t work for free, future promises of payment, exposure, likes, or “it might lead to something later.” Bouncing around general ideas in a forum discussion is one thing, but developing a proper concept, art direction, mechanics, pitch, or production-ready design is paid work.
If you already have one skin done, then I’d start by creating a few simple reskins of the same obstacle shapes.
For example, if the current version is wood, make another version as a generic rubber material dark grey or coloured rubber, slightly rounded edges, and a soft sheen so it feels bouncy rather than solid.
Another easy one would be metal cooler grey tones, sharper highlights, maybe a few scratches or bolts, so the same obstacle instantly feels heavier and harder.
That way you are not redesigning the whole game each time. You are building a small visual language where each material suggests a different behaviour.
Wood breaks. Rubber bounces. Metal is heavy. Glass shatters. Ice slides.
That gives you both the physics and the visual style to play with across different levels.
But the details are still very limited, so it is hard to give anything more specific without knowing what the game actually needs to do.