There is a real difference between Report Writer, Data Warehouse, Data Discovery and Data Scientist.
The Report Writer implies a relational database, with a 4GL language sitting on top. With SQL queries, Summary & Detailed reports,which are scheduled and consumed.
The Data Warehouse is a layer up, moving data from relational sources (ETL) to a Star / Snowflake Schema database which flattens the rows for fast consumption. Changes must be thought out in advance and with precision, which slows down the process.
Then there's the Data Discovery role, which uses tools such as Power Pivot, Qlikview or Tableau. These tools allow mashing of disparate data sources in a desktop environment. Users can see data instantly, creating graphs and charts, slicing data, etc. which could them be ported to a server for enterprise consumption.
And on top of that sits the Data Scientist. Where they mine the data. And forecast. And use statistical analysis. I like this video by Mark Tabladillo giving overview of Data Mining using Power Pivot & Excel.
Mark gave a presentation to the Tampa Bay BI group a while back but this video gives good intro to Data Mining.
So the four roles could be separate. Or they could be combined into one person. Depends on the organization and developer skill set. Either way you slice it, users want and need access to their data to base decisions.
There's also the advent of Big Data. Not sure which bucket this would be yet.