Indexes can be utilized in a regex search, it would just need to scan the index, but that is likely more performant than scanning the entire collection. For example if you have 10,000 documents in the collection, and an index on "title", but title was not unique and only had 1,000 different values, then the index would still be very helpful against the regex search since it could match against 1,000 instead of 10,000 checks.

type "double" was mentioned twice.

Bump