X vs Y - which BI tool handles large datasets better?

The real gotcha is asking which one handles large datasets better. Neither does, directly. They're both just query builders. The difference is which one lets you screw up the configuration faster.

You're asking about setup and management for a non-data person. That's how you end up with a direct API connection. Both tools will happily let you do that, and both will fall over at the same point. The easier setup is the one that leads to the harder failure.

Look past the sales page. They both rely on a proper warehouse. The "performance" question is really about which one pushes the computation back to the warehouse most consistently, and they're about equal. So now you're back to comparing a warehouse.