The meta-argument you attribute to Spinoza is closely related to the rule-following regress considered by Wittgenstein in Philosophical Investigations. To apply a rule in a particular situation we first have to interpret what it means, he muses. But then we need another rule to make the interpretation, and another, and another. We can no more apply a rule, it seems, than a runner can start running in Zeno's Dichotomy. Nonetheless, we do manage to follow rules, we read, we write, we play chess (and runners do run). Therefore, concludes Wittgenstein, "there is a way of grasping a rule which is not an interpretation".
Similarly, there is a way to "grasp" the willing without "an act of will". The meta-argument is based on the so-called volitionist theory of action, usually traced back to Descartes, and accepted by Spinoza, Leibniz, etc., before Kant. It is not an argument against free will, but one of many arguments against that theory. As confirmed by psychological studies, we do not first perform an act of will, which then causes us to do something, we just do it, willingly. This is reflected in the language, rather than say "I willed my hand to rise, and it rose" we say "I raised my hand".
Hacker's Human Nature places the theory of volition in the conceptual context of modern philosophy and science. He summarizes Wittgenstein's response to the meta-argument (Wittgenstein was influenced by Schopenhauer on the will issue) as follows (pp. 148-152):
"When one utters a sentence, every word is spoken voluntarily, but it would be ridiculous to claim that one consciously performs successive acts of will, one for each word (or phoneme?) an instant before utterance... The willing must not be conceived as doing something, the doing of which then causes the movement of one’s body. That would be a case of bringing about the movement of one’s body by doing something else. Rather, the willing would have to be an ‘immediate causing’".
The law of cause and effect, a.k.a. the principle of sufficient reason, is another postulate of traditional metaphysics, once considered impeccable but controversial today. Belief in it leads to determinism, which unlike the volitionist theory is coherent if implausible. The traditional arguments for it often confuse causes with reasons, and reasons with necessities. Philosophers started broadly questioning it after Kant, who limited it to "phenomena". According to the standard interpretations of modern physics, determinism is false, there are effects that have no causes, or are self-caused depending on terminology. Physics itself however, as empirical science, can not settle metaphysical matters. There has been a burst of interest in free will in light of recent neuroscience experiments designed to test "folk intuitions" about it. Roskies surveys their results in How Does Neuroscience Affect Our Conception of Volition?, and concludes that "to date no results have succeeded in fundamentally disrupting our commonsensical beliefs".
The modern consensus is that libertarian free will, as it is called, is coherent and unfalsifiable, but so is determinism, and the two are incompatible (there is also compatibilism which redefines the meaning of "free" in "free will"). So we have a choice to believe or to disbelieve it. As William James put it, "My first act of free will shall be to believe in free will". Information Philosopher gives a nice overview of historical and current views on free will.
