Re your 1st question Colinearity does not make the estimators biased or inconsistent, it just makes them subject to the problems Greene lists (with @whuber 's comments for clarification).

Re your 3rd question: High colinearity can exist with moderate correlations; e.g. if we have 9 iid variables and one that is the sum of the other 9, no pairwise correlation will be high but there is perfect collinearity.

Collinearity is a property of sets of independent variables, not just pairs of them.

Re your 1st question Collinearity does not make the estimators biased or inconsistent, it just makes them subject to the problems Greene lists (with @whuber 's comments for clarification).

Re your 3rd question: High collinearity can exist with moderate correlations; e.g. if we have 9 iid variables and one that is the sum of the other 9, no pairwise correlation will be high but there is perfect collinearity.

Collinearity is a property of sets of independent variables, not just pairs of them.