Remove comments w.r.t. potential db optimizations

DB queries already are performed on their own thread, so spawning a
thread to do db work is unncessary.

src/cache/disk.rs

@@ -77,24 +77,6 @@ impl DiskCache {
             db_update_channel_sender: db_tx,
         });
 
-        // maybe this is better? in theory this sounds like it should be faster
-        // but in practice I haven't seen any perf improvements
-
-        // let ret = Arc::clone(&new_self);
-        // std::thread::spawn(move || {
-        //     tokio::runtime::Builder::new_current_thread()
-        //         .enable_all()
-        //         .build()
-        //         .unwrap()
-        //         .block_on(db_listener(
-        //             Arc::clone(&new_self),
-        //             db_rx,
-        //             db_pool,
-        //             disk_max_size.get() as u64 / 20 * 19,
-        //         ))
-        // });
-        // ret
-
         tokio::spawn(db_listener(
             Arc::clone(&new_self),
             db_rx,