The only way to evaluate how useful a columnstore index is for a specific database is to compare performance with and without the columnstore index. But you have seen that it takes a lot of time to build a columnstore index, so you may be hesitant to drop the index and then have to recreate it again, just for a test. Luckily, there is a better option: adding the query hint
OPTION(IGNORE_NONCLUSTERED_COLUMNSTORE_INDEX)
to a query will instruct the optimizer to execute the query as if the nonclustered columnstore index does not exist. This hint works for nonclustered columnstore indexes in both SQL Server 2012 and SQL Server 2014. Unfortunately, there is no similar option for clustered columnstore indexes in SQL Server 2014. If you want to do performance tests for the clustered version, dropping and rebuilding the index in between tests is the only way to go unless you want to maintain two tables, one with the clustered columnstore index and one without.