While I absolutely agree with this, and think Twitter does things that make it worse: no automated (or human-driven in bulk) abuse-detector will ever work perfectly. People are creative. Some amount of manual blocking is a reasonable expectation... just not as much as is necessary now :)