From your production function: $f(K,L)= 1.8(K^{0.6} + L^{0.6})^{2} \Rightarrow f(K,L) = 1.8(K^{1.2} + 2(KL)^{1.2} + L^{1.2})$
You are able to see that $f(\lambda K, \lambda L) \Rightarrow \lambda^{1.2} f(K,L)$. Say you double both your inputs. Increasing returns to scale tells you that your output will be more than doubled, which is precisely what $\lambda^{1.2}$ represents. The level of $K$ and $L$ does not affect this conclusion. The point where the level of $K$ and $L$ play a "role" is that for levels such that $K + L >1$ each unit of input results in more than one unit of output (i.e.$f(K,L) > 1$) (again note that we are talking about levels here and not changes, which would regard the returns to scale). You can easily see this by plugging in different values into your formula. But, again, this is not to do with the returns to scale of the function so it doesn't pertain the question you are asked.