I'm thinking it would've been very likely that the "4" private key would've been discovered already. There's been several incidents in the past where very poor RNGs were used for generating keys. (Debian's stock i386 kernel for like 5 years only had uint32-a few billion possible values..)