The slow performance you're experiencing is a known bottleneck where
Table.ExpandTableColumn triggers a separate API call for every single row in your primary list.
For 25,000 rows, this means 25,000+ individual HTTP requests, which is why it feels like it's crawling.
The most efficient remedy is often a Merge (Join) operation combined with Buffering to force the expansion to happen in-memory rather than via repeated calls.
Recommended Solutions
*-> Classic Table Join (Merge): Instead of expanding the nested record directly, load both SharePoint lists as separate queries. Perform a Merge Queries operation in Power BI using a common key. This reduces the workload to just two API calls-one for each list-and performs the "expansion" locally in your computer's memory.
Table.Buffer: If you use a custom function or a merge, wrap your secondary table in Table.Buffer(SecondaryTable). This "freezes" the data in memory, preventing Power Query from re-evaluating the entire source for every row it processes.
SharePoint REST API with OData: For maximum speed, use the SharePoint REST API to fetch data in batches (e.g., 5,000 items at a time). You can use $select to pull only the columns you need and $expand to handle lookups more efficiently than the standard connector.
Filter Early: Ensure you are filtering rows and removing unnecessary columns as the very first steps in both queries to minimize the data payload before the join occurs.
Reference:
https://learn.microsoft.com/en-us/power-query/optimize-expanding-table-columns