In the case of the Helios accident (the final report for which I have right in front of me), there were no passengers entering the cockpit, and there was apparently no attempt to revive the pilots.
What actually happened, as far as the accident board could piece together, is that a flight attendant entered the cockpit just before the first engine flamed out from lack of fuel, and sat in the captain's seat and appeared to be attempting to fly the airplane. This attendant, the only male member of the cabin crew (and thus easily identified), also held a valid U.K. CPL and a valid first-class medical.
CVR evidence indicates that standard emergency procedures were used to access the flight deck, but the board was puzzled by the fact that this FA waited until 30 seconds before the engine flameout to actually open the door. A possibility is that he didn't know the proper access code and could not actually open the door until power to the locking mechanism was lost as engine failure occurred. Only the Cabin Chief normally had the access code at Helios.
Once the door was open, the FA tried to sit in the left seat and appeared to be trying to fly. He didn't see the F-16 escort at first. At one point he sat up and remained motionless, as if unconscious. After the engines flamed out and the aircraft began to descend, as it passed through 7000 feet or so the FA appeared to see the F-16s. In response to their signals to follow them to the airport, the FA merely pointed down (as if to say that he had no choice but to go down now). Shortly thereafter the aircraft crashed. Nobody else was seen moving on the flight deck. The CVR recorded two weak attempts at a mayday call, but this was not on the radio record; perhaps the FA was too far gone to remember to key the microphone.
The CVR evidence seems to indicate an entry onto the flight deck using standard procedures, but that doesn't explain why the FA waited so long to enter (he waited hours to do so, in a depressurized airplane with an unresponsive flight deck), nor does it explain the remarkable coincidence of him opening the door only seconds before the first engine flameout. This does tend to encourage speculation that he was unable to open the door for some reason and gained access only after it unlocked on its own, by which time it was too late. The report concluded that the secured cockpit door was not necessarily a contributing factor, but there isn't enough information to know that for sure. There were some anomalies in flight-deck access procedures at the airline, but those were not necessarily contributing factors, either.
The bright side to the story is that the flight deck was thoroughly protected against any evil terrorists on board for essentially the entire flight.
There are far more accidents than terrorist attacks aboard aircraft, so it seems to me that the emphasis should be on dealing with the former rather than dealing with the latter … which means that access to the flight deck should be made easy in any emergency situation that cannot be readily provoked by passengers. If the flight deck is unresponsive, the risk of letting terrorists enter the flight deck pales to insignificance (in fact, the risk of terrorism is already vanishingly small even in normal flight), so the flight deck door should unlock in any emergency. It is not sufficient to assume that the right person in the cabin crew will be able to open the door.
There is too much paranoia in the aviation industry today in relation to the very small threat from the bad guys, whereas the standard paranoia about safety—which is far more justified—seems to be waning. This is a worrisome trend. One need only look at the numbers to see where the real concern and emphasis should be placed, and it's not on dealing with mostly imaginary bogeymen.