Weighted By Taste, Not By Genre

A neighbourhood recommender predicts user $u$'s rating for item $i$ as the similarity-weighted average of ratings from users who already rated it:

$\hat{r}${ui} = \frac{\sum{v} s_{uv} , r_{vi}}{\sum_{v} s_{uv}}

Three neighbours have rated the film Dune:

neighbour	similarity to $u$	rating of Dune
A	0.8	5
B	0.4	2
C	0.2	1

What does the model predict on the 1 to 5 scale?