Has anyone recently deployed to Reltio for master and/or reference data management?  Any lessons learned to share?  

I have never used Reltio, but there are general principles that are worth following when changing the data management model.

Clearly define your objectives and goals for moving to a cloud-based MDM model. Understand what you aim to achieve, such as improved scalability, accessibility, or cost savings. Conduct a thorough assessment of your current data architecture and management practices. Understand the structure, quality, and relationships within your data. Finally ensure seamless integration with other cloud-based or on-premises systems. APIs and connectors should be well-designed to facilitate smooth data flow between systems.

Hello, I haven't a direct experience with Reltio but with pleasure I dedicated time to look around relevant information about the topic.
There is an interesting post on the Reltio community
https://community.reltio.com/blogs/chris-detzel1/2022/05/19/master-data-management-modernization-empire-life

In the short, it is necessary a substantial job activity shifting from the traditional relational data model to Reltio's model. This requires reworking data mappings and ingestion processes to conform to the new model.
Of couse as said above it involves a substantial effort but also an opportunity to uncover and rectify design flaws in the original solution. 

Regarding the lessons learned, looking at the materials online,I believe the emphasis should be on adopting a different approach with solutions like Reltio, where Master Data Management (MDM) is seen as a practice of continuous evolution.