If it helps any, I have personally avoided table partitioning for many years even in Standard edition with very large and wide tables in the hundreds-of-millions. Proper indexing, understanding my users, dedicated I/O, proper placement, etc have helped me excel with great performance with very little resources.

Now that my tables are getting to 4 billion records, other options needed to be considered, especially with loaded backdated data into a billion record table clustered on time and users making huge reads with their queries. Partitioned views and table partitioning became that option taken to help in those situations, but not just for increased performance.

However, keep in mind, in my case this is a DW (OLAP) environment and not transactional (OLTP).