There is at least one good consequentialist reason for punishing attempted murder less severely than murder. If the two crimes are punished equally, then the law will not deter someone who has tried and failed to murder from trying again!

To the extent that the "successful murderer" is simply luckier than the "failed murderer," your question raises the vexing problem of moral luck. Consider the following two cases:

1. John drives after drinking way too much at lunch and is pulled over almost immediately and arrested.

2. Jack drives after drinking way too much at lunch and almost immediately runs over and kills a family of four crossing the street.

It seems that there is no morally relevant difference between these two cases. That is, the only difference between them seems to be that John was lucky enough to have been pulled over before he could cause any serious harm. Yet (at least in our current criminal justice system), we punish Jack much more severely than we do John. John is charged with driving under the influence, but Jack is guilty of vehicular manslaughter. Should John receive more lenient treatment merely because he was lucky?

Two of the best-known essays on moral luck are by Bernard Williams and Thomas Nagel. You can find them here and here, respectively. There's also a nice (and free) overview in the online Stanford Encyclopedia of Philosophy.