The only possible difference between using ARs versus MVCP/MVCS is that there is the overhead of AR translation. However, if you're limited to only a single source and destination address space, then this will happen once, and the results will be cached in the ALB; so the issue is probably moot.
What you might want to consider is the cost of totally trashing your data cache with all of stuff you're copying. To minimize that cache trashing, read up on the special pad characters for MVCL and MVCLE. When these instructions are not performing padding (which would be the case for an address-space copy), pad character X'B0' indicates that there is no immediate intent to use the data, thus the cache can be bypassed.
