Once upon a time, I had this question come up during an onsite visit: someone wanted to create a ‘materialized’ copy of the remote data locally as an indexed view. I had never heard that requirement before, but since the motivation was genuine, I decided to validate it.
There is a long list of what is permitted (and what is not) in indexed views. One fundamental requirement is that the view on which the index is created, must be schema-bound. And this fundamental requirement, actually nails the answer straight away: any attempt to create a schema-bound view which references a remote table ends up with the SQL error 2014:
Remote access is not allowed from within a schema-bound object
Now that it is clear that such ‘projection’ is not possible, we have to step back and understand the real problems why this requirement was posed. In most cases, such ideas come up for improving performance of distributed queries (DQ). For all such cases, I suggest you first start by reviewing the excellent talk from Conor on DQ.
But in many cases, especially where multiple remote data sources are involved, probably the best way to optimize the queries is to consolidate the data inside a data mart or warehouse, and then possibly consider the columnstore index (or the new Clustered Columnstore in SQL 2014) feature to optimize the aggregation-type queries which are very typical in those scenarios.