You didn't mention scalability or elasticity. How does this bespoke stack react when demand spikes to 200 or 2000 concurrent users? Can it react instantly in real time, or does that require some admin to manually add resources?

Are you striping your data and compute clusters across all three AZ's in the region, with automatic failover?

What are you using for data governance? Can you mask PII/PHI or tag sensitive data elements? Is there any data lineage tracking built-in? Can you share your data in real-time with other users - even if they're on AWS or Azure? How about a public Data Marketplace for monetizing your data assets? How will you build silent client failover in the case of a region or cloud outage?

If your view of Snowflake stops at "cloud data warehouse", you're missing out on a lot of the value-add conversation.