if this is true, then would it be a better investment to have the 2nd team produce a fuzz testing/systematic testing mechanism instead of producing a secondary copy of the same system?

In fact, make it adversarial testing such that this team is rewarded (may be financially) if mistakes or problems are found from the 1st team's program.