(or, while final cause can be best for your map, efficient cause is the primary out in the territory)
Describing a phenomenon in terms of final cause is often the most useful and effective way to explain the given phenomenon for one's purposes. For example if you want to know why a plane flies or why a computer program operates the way it does, it's because it was designed that way. A squirrel climbs trees because it wants to eat nuts, it wants to eat nuts because it wants to live, it wants to live and reproduce because evolution designed it that way. Evolution designs organisms a certain way because it wants to maximize genetic fitness. A person acts a certain way because they desire the expected outcome.
It's virtually never a good answer to explain a plane's behavior in terms of the atomic and subatomic interactions which ultimately account for all the efficient causes behind the plane's behavior (except possibly in extremely advanced military fighter or space shuttle research laboratories or something).
However, in every case of final cause we observe, science at some point over the last two and a half millennia has found corresponding efficient causes. And, more importantly than finding that these efficient causes correspond with final causes, science has found that the efficient causes are *primary*. Without legs, the squirrel won't climb a tree, no matter how much it wants the nuts. If you take away the necessary brain function, the free will disappears. Without reproducing species and the rest of evolutionary mechanics discovered by science, evolution won't go on evolving things.
But, I would not go on to say that final cause, freewill, experience, and so on are illusory and "all that exists is efficient cause". When someone describes behavior in terms of final causes, or describes experience or free will, in the terms and meanings they are using, all of those things certainly do exist. You could no more deny final cause than to deny efficient cause - because ultimately, the final causes we observe and talk about, we have found they DO have corresponding efficient causes.
It's just important to remember that while the final cause is often epistemologically primary, so to speak, the efficient cause is metaphysically primary.
(this is just another way of trying to help dissolve the general classes of reductionism, freewill/determinism, and qualia issues - most often these are the result of metaphysics/epistemology confusions, or in LessWrong parlance, map/territory confusions.
the advantage of thinking of it this way is to try to see a more general relation between final cause and efficient cause that applies not just to mysterious brains and minds, but to much less mysterious events like squirrels legs climbing trees. when you have a clear idea of why reductionism/compatibilism is obvious in non-mysterious contexts, it's much easier to see that it applies just as well even in the mysterious contexts).