In science, everything is "data", "hypothesis" or "theory", and models are the mechanism scientists use to link "data" and "theory" - this is as true for Climate Science as it is for Physics or Chemistry.
The problem with the icecaps is, at least in part, one of chaos - it can be compared to knowing with 99% certainty that an earthquake will occur near San Francisco in the next few hundred years, but not being able to tell you at what date & time it will occur.
The problem is that in the climate debate, some people then pretend that, because we cannot predict the exact date & time of the earthquake, we know nothing about earthquakes and we are wrong in predicting that the earthquake will occur.
The Earth's climate is indeed a complex, nonlinear system, but it is not entirely without constraints, many of those constraints have been figured out, often from the basic physics. You may not "think" that people are clever enough to understand it, but there are some *very* clever people working on this.
This is also why no serious climate scientist is claiming that in 2100 the average temperature will have gone up 2.7548C, what they are claiming is that, given a few assumptions on the evolution of the rise in CO2, the most likely result in 2100 will be a rise in average temperature of between 2C and 4.5C
Further, the uncertainty is much bigger on the high end than on the low end. We are virtually certain that, unless significant reductions in CO2-output are realised, the rise will be no smaller than 2C, and the most likely figure looks to be 2.5C-3.5C.
The science has come far enough that the biggest uncertainty left isn't in the science, but in the human factor, which is the most important bit that we cannot integrate into physical models.
And to give even more food for thought: the temperatures during an ice age are estimated to have been, on average, about 5C lower than today. So while 2C may seem like a small amount, realise that a 5C drop in average temperature caused the ice-caps to cover most of Canada & Western Europe, and sea levels to drop by almost 100 meters.