questions about Change Data Capture

Hi.  We're going to take a quick look at this but wonder about the following before we do:

1) will it work if recovery is simple on the db in question?

2) will it work on 2008 std        Non R2?

3) can the overhead we'll experience (say in extra run times during high volume inserts/changes/deletes) be somehow quantified or predicted? 

4) if we want the target of the cdc data to be on a different server and we dont use sql's linked server feature, will we need to keep the captured data on the source server?