Starting SQL Server 2017 there is no need for Clustering or Listener to deliver a solution for your scenario.
Few things to consider though:
- AlwaysOn enables READ_COMMITED_SNAPSHOT isolation level on a primary server. This means overhead on TEMPDB and extra 14 bytes per row on every row change
- In the case of Asynchronous Mode, the data recency on a secondary server can be close to the primary server.
- Versions older than SQL Server 2017 require WSFC.
Therefore, AlwaysOn AG readable secondaries have pros and cons in comparison with log-shipping:
- Pros:
- No need to interrupt connections because no need to restore logs
- Data can have a nearly real-time recency
- Cons:
- Enterprise Edition only
- 14 bytes overhead per changed row on the primary replica, therefore consider to change fillfactor from 100 to 90 to avoid page splits overhead
- Way harder to maintain
Regarding your question:
would anyone know if this is a good or terrible idea?
AG readable secondaries are definitely worth of POC trial, especially if your company has required skillset/resources
(Disclaimer: this post is based on my personal opinion)