[bjt]

I'd add the biggest risk of over-engineering: Building the wrong thing. Very often, the needs we anticipate end up not being realized.