Neither really makes sense from a statistical perspective: your number is effectively the true value, and there is no repeatable measurement process you’re evaluating.

Your use of the non-statistical meaning of “accurate” as “free from error” is spot on.