The first two lines are fine, and there's nothing wrong with the equation $0=0$ (or any equation equivalent to it). The wrong step occurs in going from the second to the third line.
This step is, from its structure, the argument that if $ac=bc$ then $a=b$. However this rule does not apply for $c=0$. It does apply for other numbers because those are invertible, that is, there exists a value $c^{-1}$ so that $cc^{-1}=1$. Now there is a general rule that if $x=y$ then $xz=yz$ (and that rule BTW also holds for $z=0$), and therefore for invertible $c$ you can conclude from $ac=bc$ that $acc^{-1}=bcc^{-1}$, that is, $a=b$. However for $0$ this derivation doesn't hold because there's no number $d$ such that $0d=1$, and indeed there are many counter examples showing that for $c=0$ this rule doesn't hold, like $2\cdot 0=3\cdot 0$ but $2\neq 3$, or indeed your fake-proof, where the fact that $c=0$ is (slightly) obscured by writing is as $(2-2)$.