Schubert came extremely close to discovering the JSJ-decomposition of 3-manifolds in his paper "Knoten und Vollringe" (1953). With a little more work, one could turn Schubert's paper at least into something equivalent to the JSJ-decomposition applied to knot and link complements in $S^3$. That would at least have allowed people to conjecture the JSJ-decomposition for 3-manifolds, around 20 years earlier than it was.
Schubert came extremely close to discovering the JSJ-decomposition of 3-manifolds in his paper "Knoten und Vollringe" (1953). With a little more work, one could turn Schubert's paper at least into something equivalent to the JSJ-decomposition applied to knot and link complements in $S^3$. That would at least have allowed people to conjecture the JSJ-decomposition for 3-manifolds, around 20 years earlier than it was.