According to: https://docs.microsoft.com/en-us/azure/databricks/integrations/bi/power-bi yes. It is standard sql so should be not problem. However when querying this way there can be issues when querying by field which is not used to partitioning or z-order. Personally when model is not extremely big I prefer to load everything to PowerBI and set daily reload. When it is real time data is better to use streaming (Databricks stream sink to Azure event hubs connected to Azure streams which is consumed by PowerBI).
