Not necessarily. A working simulation (for some testable subset of states) doesn't carry any hard and fast logical implications about your understanding.
On the other hand, assuming no errors in implementation then a broken simulation which you had expected to work directly implies that your understanding is flawed.
On the other hand, assuming no errors in implementation then a broken simulation which you had expected to work directly implies that your understanding is flawed.