You need to consider all of the markers tested. The fact that others did not mutate is also important. Note that the rates are quoted per meiosis (transmission event). There are two legs when counting generations, so the number of meioses is twice the number of generations in FTDNA's model.
Note that the calculation gives ranges that are too wide to be used to pinpoint the time when the common ancestor lived. It is instead used to determine the chances of finding the common ancestor within a certain time period if you worked backward. For example, there is a 50/50 chance of finding him in 7 generations, although my calculator uses different assumptions from FTDNATiP.
Of course, this does not have much practical value, if you have a brick wall, since you cannot work backward. It is mainly useful for determining whether 2 people are related. Howeever, you can simply use FTDNA's guidelines. I think they are on your web page under "Understanding Your Results."
Generally, we quote generations, but a good number to use is 33 1/3 years for the generational interval. it is different from the one used in genealogy, which is the parent's age at which the first child was born. This is the average of the father's ages when each son was born. During the times for which we have records, it is somewhere in the mid 30s.