« At least almost all Collatz orbits attain bounded values, and other significant corollaries on the Syracuse problem »