Don't you think that the scientists who literally dedicate their life to understanding the climate are aware of the difficulties and have developed methods to get useful predictions out of ensembles of models? The predictions used for example in the IPCC reports get their error bars at least in part from the disagreement of the different models.

Regarding sea ice: I think a temporary increase is not unexpected. As ice melts the salinity of the ocean goes down, which makes it freeze a bit easier.