What's the goal? If aggregation itself is the goal, I mean sure. But if truth is the goal, then at a minimum you will want to:
1. Rate pollsters based on some sort of evaluative methodology -- for instance, ensuring that they are in fact taking polls.
2. Attempt to assess the partisan lean of a pollster when evaluating the poll
3. Have some mechanism to prevent your poll aggregation from being overrun by junk
4. Use weights of some sort to give more emphasis to more important polls.
RCP is akin to judging basketball teams by their W/L record alone, without any adjustments for margin of victory or strength of schedule. If that's your cup of tea, go for it.