I think it's not really the size of the project per se but the proportion of domain complexity. For problems with low domain complexity but involve a lot of bo

I think it's not really the size of the project per se but the proportion of domain complexity. For problems with low domain complexity but involve a lot of boiler plate code, a more expressive language wins out. Likewise, if you spend the majority of your time thinking about the problem and little time on the actual implementation, then the language may not be your bottleneck.