Over-engineering is bad. Period. It's costly and introduces complexity that you and the customers don't want to deal with.

On the other end, just because you think you cannot predict the future 100% correct, it doesn't give you an excuse to put no effort to do it. You don't ship into the ocean without a map.

Any company would experience some growing pains if it succeeds. As architects, our job is to do our best to predict the future and design an blue print that meets that predicted needs :)