I'm not sure having a complete Unicode v11 font set of 137,000 characters would be the best use of resources. However, I have to agree that limiting support to use latin characters is disappointing. Given that Google, Red Hat and the GNU project have released more complete font sets under licensing terms similar to library code baked into Roku OS. For example, Google Roboto is under an Apache license, has a complete Unicode v7 character set, is less than 200kb as a TTF and is designed to still look good at a wide range of resolutions including budget phones that are below 720p quality.
I understand there are design constraints to creating an embedded device but in this case I am having a hard time coming up with a good excuse why cyrillic isn't standard in Roku OS and would be left up to being on a per channel basis for support.