With voters concerned about McCain's age and his history of skin cancer, it remains to be seen whether they will be reassured by a woman with little to no experience on the federal stage who is a heartbeat away from the presidency.
Let us remember, however, that half of all presidents who have died in office have died from assassination, not natural causes.
If we take the past data as indicative of the current odds of death while in office (not a safe assumption given the small sample size and the time differences between sampling, though it's the best information available), then the average president could be said to have a 9.3% chance of dying by assassination while in office and a 9.3% chance of dying from natural causes, or an 18.6% chance overall. If we take John McCain to be 50% more likely than average to die from natural causes (probably an overestimate), then his odds of dying in office would be 23.25%, a less than 5% difference.
As I admitted, this statistical data is unreliable. Perhaps fewer than 50% of presidents who die in office in the long run die from assassination. It is possible that the real figure may be more than 50%. However, it is clear that assassins kill a significant number of presidents. At best, health and age combined have only a fractional impact on the likelihood of a president dying in office. In the other instances, Barack Obama is no safer than John McCain. So while Sarah Palin's experience is a valid issue, it should not be much more significant than that of any other VP candidate.
Update: There's an examination of a similar topic in the WSJ Best of the Web. They kind of flip the topic on its head, though, and compare the likelihood of the two more experienced candidates serving as president by the end of the term (McCain versus Biden). They don't account for the high risk of assassination that all presidents face, however, which should bring both McCain and Obama's chances of death in office closer to the mean than if only considering the SSA data.