buffer cache and proc cache are separate so I don't think it would affect PLE directly. Are you actually seeing any memory related error messages?

with such a large CACHESTORE_SQLCP you should test with sp_configure option 'optimise for ad-hoc workloads' turned on, this will reduce bloat.

Also SP1 is way out of date, you should patch to SP4, there may be improvements\fixes that will help