In some cases we can achieve sustainability by taking small steps, and in others by giant leaps. It is up to bright people to find the (societal) cost benefit optimum in each case. They should account for things like opportunity costs, lock-in effects / sunk costs, and feasibility.
Sustainable development is in small steps and giant leaps.
For example, in the built environment we deal with situations where we need steps and others where we need leaps. We need steps in the existing building stock, because demolishing and rebuilding everything is unsustainable. We, at the same time, need leaps in the construction of new buildings to prevent us from getting locked up in soon-to-be-old technology. The same goes for other initiatives: some work best by incrementally changing behavior, while others need a launching customer to get off the ground.
What about the (needed) exponential growth of installed sustainable solutions? Both steps and leaps contribute to that. Generally speaking, small steps are taken when there is potential for many many steps. Many small steps together makes for exponential growth just as much as the fewer giant leaps in the most advanced parts of our society.