That episode (E2) has a 9.4, tied for the most of any episode of this series, and most people seem to agree that it's one of the best of the series, if not the best, so the rating seems to match the quality. Also, IMDb already gives less weight to low ratings. For example, the unweighted average for the current episode is 6.7, whereas the official average is 7.3. I just ran the numbers and 7.3 is exactly what you get when you replace all of the 1-3/10 ratings with 4/10 ones, so they're basically already doing a 4-10 scale, like you want.
As I just pointed out, the 1/10 ratings aren't impacting the average any more than if they were all 4/10s. What are skewing the average, though, are all of the 9/10 and 10/10 ratings. If we really care about fair, unbiased ratings and feel that these episodes are only 5.5/10 quality, we should have a bigger problem with all of those really high ratings that do a lot more to skew the average than the really low ratings that are barely even being counted.