But you can't say it's an accurate number if not all reviews have scores.
One F score can drastically effect an average rating, likewise an A+.
When the sample size gets large enough, yes you can. Unless there's some sort of bias to the reviews that don't have scores.
That's statistics.