Node.js has some very good bindings for SQLite thanks to the SQLite3 package, but unfortunately, since Node runs single-threaded, all queries are done in the same connection. SQLite runs all queries per connection in series, meaning that it is impossible to actually do parallel queries in SQLite under Node. For more details check this https://github.com/mapbox/node-sqlite3/issues/394
This is true whether you wrap your query-logic in async.each/async.parallel, or any of the other helper packages for parallelising/serialising database IO. At the end of the day, to benefit from Node's async IO architecture under SQLite3, you are going to need more than one processing thread.
How can this be done? :)