What was done
The author used a comprehensive set of collocated surface and satellite observations to calculate the amount of solar radiation absorbed in the atmosphere over equatorial Africa, after which these empirical results were compared with the theoretical predictions of three different general circulation models of the atmosphere.

What was learned
It was found that the climate models predicted an atmospheric solar radiation absorption that was much too small.  The models did not properly account for spatial and temporal variations in atmospheric aerosol concentrations, creating regional and seasonal underestimation biases of up to 30 Wm^-2.  On top of this problem, it was found that the models likely underestimated the amount of solar radiation absorption by water vapor.  And on top of this problem, it was found that the models may also be underestimating the amount of solar radiation absorption by clouds.  And on top of this problem ....  Well, you get the picture.

What it means
With such inadequacies in one of the basic features of all climate models, one would think that one would want to be very cautious in interpreting the predictions of such models.  At the very least, one would think - if one is a thinking person, that is - that one would not want to reconfigure the entire economic/energy structure of the planet without something more substantial to go on than these demonstrably inadequate predictions.