Reichler and Kim Say Reliability of Computer Climate Models Cannot Be Known

Who says NASA is biased in favor of the anthropogenic global warming hypothesis? They posted this among their "media alerts":

Although model-based projections of future climate are now more credible than ever before, the authors note they have no way to say exactly how reliable those projections are. There are simply too many unknowns involved in the future evolution of climate, such as how much humans will curb their future greenhouse gas emissions.

How can something which is of unknown reliability be more credible?

Apparently, the paper which is the subject of this particular "alert" concludes that current computer models do a pretty good job of simulating today's climate:

In the study, co-authors Thomas Reichler and Junsu Kim from the Department of Meteorology at the University of Utah investigate how well climate models actually do their job in simulating climate. To this end, they compare the output of the models against observations for present climate.

So, the argument is: if they can simulate today's climate, then they are "credible" as projections of climate conditions in the years to come.

True, but that bit about unknown reliability is worrisome, isn't it?

The example given doesn't dispel the worry, since the amount of change in future emissions is usually assumed for different scenarios in order to show what the projected result would be for each of them. What are the other unknowns?

And, to what extent have the current models been adjusted to get them to simulate observed climate conditions? Have they been "tweaked" to the point that they no longer follow known physical principles? (Miskolczi argues that the models do not follow physical laws.)