Multiple consistent dates from a site, or from a burial, provide what statisticians call redundancy — if all the dates cluster around the same value, you have more confidence that you are measuring what you think you are measuring.

A single date provides no such redundancy.

If something went wrong with that one sample — if the sample was contaminated, if the chemistry was affected by postdepositional processes, if there was something unusual about the carbon in the material being dated — you have no way of knowing, because you have nothing to compare it to.

The researchers who published the 1998 report were aware of this limitation.

They noted it explicitly.