Query folding creates unusable queries in merge

It is not uncommon for Power BI queries to experience performance issues when joining large tables. The additional clauses added to the ON statements in your query are causing the issue, and removing them has improved performance. These clauses are used to handle null values, but they can be problematic in certain situations.

One solution to improve performance is to use native SQL queries instead of relying on the Power Query Editor. This would give you greater control over the query and allow you to optimize it for performance.

If you prefer to continue using the Power Query Editor, there are some potential workarounds. For example, you could try breaking down the query into smaller, more manageable steps or use other techniques such as filtering or sorting the data to improve query performance. Additionally, you could try modifying the Snowflake settings to optimize query performance, such as increasing the warehouse size or adjusting the concurrency level.

Ultimately, the best approach may depend on the specific needs and constraints of your project. It may be helpful to consult with a Snowflake expert or a Power BI consultant to find the best solution for your particular situation.