It appears this problem isn't specific to extracting data. It happens whenever I try an operation on a BigQuery datasource with more than 1 table.
It generates SQL like:
SELECT [tableA.rl_gender] AS [none_rl_gender_nk] FROM [tableA] INNER JOIN [tableB] ON ([tableB.email] = [tableA.email]) GROUP BY 1
With the error:
[Simba][Gauss] (20) Query execution failed. Table name cannot be resolved: dataset name is missing.
The original BigQuery connector in Tableau 8.0 used a different syntax which may not be able to express what you need to do. As an experiment, can you please try to create a new BigQuery connection in Tableau 8.1 and determine if the multiple-table option still fails? If so, please report this to our Support team (email@example.com) so they may work with you to collect log files and more details about the problem.
If you find that the new connection in Tableau 8.1 works, you can use the Replace Data Source option under the Data menu to switch your worksheets to use the new connection. While this won't preserve many important aspects of your original data source (such as calculated fields, captions, aliases, etc.), it may help unblock you from your current challenge. If you're comfortable editing the XML of your Tableau workbooks and data sources, you can change the BigQuery connection attribute from connection-dialect='simba-sql' to connection-dialect='google-bql'. If you don't have such an attribute in the <connection> tag, consider creating one.
I hope this helps,