The difference between the reports reflects a difference in counting methods. While comScore tallied unique visitors, or the number of individual people who visited the two sites in a given month, SimilarWeb counted the number of visits themselves. Put another way, if you visited Google every day in July, comScore would count you as one visitor and SimilarWeb would count you as 31 views. The comScore report also didn't include each company's various properties, although Google boasts more pageviews than Yahoo without them, too.
That difference is significant, since the web advertising industry generally runs on visits, not visitors. It also theoretically demonstrates a certain lack of commitment among Yahoo users. Google has fewer unique visitors, but each one returns to its products frequently. In contrast, Yahoo has more unique visitors but they return less frequently.
Perhaps more importantly, though, this is all a cautionary tale on the squishiness of web analytics, which each research firm obtains, parses and interprets differently. There are no firm numbers in this space, even when comparing the same indicators -- so when someone says Yahoo has “more traffic” than Google, it’s very rarely that simple. Sorry, Yahoo.