It's sufficient, efficient, but has chances of being too optimized. You could land on all sorts of other entries on the way:

[Nikers et al, 1983] , [NH4] etc

I know, I know, it's unlikely. But being a programmer, I tend to cover my bases. I'd rather waste afew milliseconds typing the extra 2 letters, than having the inconvenience of finding unecessary matches on the way.

This of course is not perfect, as I could just type [new] as much as I [new] want just to [new] throw all of [new] this off.

There is [no] ideal solution, but I certainly will [not] settle for a [non]-complete ma[nner] to look for [novel] entries.