Anyone familiar with running Scribe ETL for interfacing with GP/Dynamics/CRM is probably familiar with how challenging it can be to tune.  Running a profiler trace against it doesn't help much, since the sqltext is most often encrypted - so you really don't know what to tune.  Even in the most latest versions of Scribe it is possible to perform bulk operations, however all of your steps in the DTS must be insert operations.  Another handcuff.  What i've found is that it is important to look at the working structures you have created in GP/Dynamics or CRM and make sure they are tuned correctly.  If your DTS solution includes scripts and maintenance operations, make sure that your underlying tables are indexed or created properly. By simply turning the heap table into a clustered index can make all the difference.  For more information - contact us!