‘Model agreement is only significant if models are sufficiently independent from one another. Climate science community has mostly ignored crucial problem of model independence’