Most of our product wound up eventually eliminating the soft tooling phase and going straight to steel. It was found to be more cost and time efficient over the number of projects running to go to steel, and pay for the modifications - even if it meant cutting a new core/cavity or adding inserts than it was to run a soft tool and then re-tool the final part. It sometimes lead to very constrained design decisions (IE you can only modify the part so much in a tool safe direction before you hit a cooling line or some existing feature) but at scale it had a substantial cost and lead time reduction.