Sorry for the cold shower, but the answer is: 'not exactly'. P2V has a downtime of a few minutes at best.
Of course you must use synchronization. The caveat is that the important services must be stopped before the conversion completes.
The following steps will be executed at the end of the conversion: stop services on source machine, perform a the final catch-up synchronization, shutdown the source machine, power on the destination VM.
You should use scheduled synchronization (actually I strongly recommend you do). Do it *without* finalizing the first time. I want to emphasize this, it is very important - WITHOUT FINALIZING! You can do the bulk of cloning (full cloning) and one synchronization step. Thus you will have an idea about the downtime (it should be roughly the duration of the synchronization). If there happens to be too much to clone during the first synchronization, do another one right after that. Keep in mind stopping the services will not happen during non-final synchronizations; this will consume some more time.
What is more important - after the synchronization completes, try to power on the destination VM and ensure it works (sometimes BSODs happen after cloning, you wouldn't want this to happen to your production server, do you).
After you know it works and have an idea of the delay, you can start the final synchronization. I'd recommend you warn the users, as it won't happen unnoticed.
I also recommend you do the testing for powering on of the destination VM for all servers one by one, don't rely the other will be fine because the first one was OK.
HTH,
Plamen
P.S. Although it might be theoretically possible to do a lightning fast P2V a la vMotion, I know of no such technology currently implemented.