A student, explaining why abstract models are useful, makes a very nice analogy. If you want to go from one campus of the Berlin School of Economics to the other (by car or public transport), then you need not a description of the world as it is but a description of the world in an abstract form that allows you to get there. Below you find two ‘model’, one real and one abstract. Which one would you choose?
So, yes, models are abstract, but we use them all the time. The same goes for macroeconomics. It is not necessary to understand and know how every entity behaves. You can use abstractions and you will be fine. Income (GDP) equals consumption, investment, government spending and net exports, and maybe this is all you need to figure out how to go from Weak economy to Strong economy. Or, you add private saving and taxes and your identity reads (change in net private saving) + (change in net government ‘saving’) = (inverted change in net foreign saving). Will I miss things? Sure, but just as I don’t care what houses look like that I pass on my way from A to B I can safely ignore the motivations of single entities when I go from unemployment equilibrium to full employment equilibrium.
Nevertheless, there are more questions than the one I asked above and that means there should be more and different maps.