Have you looked at sp_setnetname?

https://msdn.microsoft.com/en-us/library/ms190318.aspx

You could do this in powershell and use Export-Csv.

+1

Merge replication uses GUID columns to keep track of unique rows.

So the delete is using the GUID value in the where clause.

The dependency on the GUID also makes it so...

You are running into the two hop problem.

This is why most linked server connections use a SQL auth login.

If the issue is a growing error log, putting a database in simple recovery model will...

Also there are many table diff tools out there. I know Redgate has one.

You could just use snapshot replication, scheduled once a day.

Tables on target instance would not be usable during load.

Another option would be change data capture and load modified data once...

How often does the refresh need to run?

How big are the tables?

What are the availability requirements during the "refresh"?

What are the different versions of the two instances?

What specifically are you skeptical about? You question is a bit vague

Recently I have been doing a proof-of-concept implementation of Apache Kafka.

I am very impressed with it performance and simplicity.

I would say monitor the fragmentation and keep doing reorg.

Then look at the rebuild options, in particular parallelism.

Also when you run rebuild, make sure your data files and log files...

Same here

Books online suggest reorg if fragmentation is < 28%

http://msdn.microsoft.com/en-us/library/ms189858.aspx

Still you have not indicated the level of fragmentation of the tables.

What is the % fragmentation?

Do you have a test system on which you can test rebuilding and see how...