My question would be, this similarity between the modeled and an asynchronous observation is related directly to the parameterized model fluxes in atmospheric and sea ice? 

I have worked with an asynchronous coupling between an atmospheric model and a sea ice model for the Antarctica region, the results are very good, capturing the seasonal and interannual variability by comparing with data from SMMR / SSMI. This coupling is different to the couplings used. 

The flux coupler was intended to gather state variables from the atmospheric component models, compute fluxes at the model interfaces, and return these fluxes to the component models for use in the next integration period, maintaining conservation of momentum, heat and freshwater.