One of the most important features for me is the ability to see how common certain phrases are (estimates are fine), but currently it only displays the number of results that get shown to me. For example, if I do exact search for "ゲーム" (Japanese for "game"), then it says there are 11 results, but if I search "の烙印を押される" (Japanese for "to be branded as") then I get 13 results, despite the two phrases obviously being nowhere near similar in frequency of usage.
To be clear, fuzzy estimates (like Google does) would be totally fine, but I just need something that can function like this.
This usage is fairly common among language learners and translators
eg:
To implement it, simply showing an "estimated number of appearance of the phrase" when doing exact search would be sufficient.
