The only problem with converting to text is a scaling issue if that text layer is nested. Hope it takes care of it together.

This is much required as mostly clients are complaining about blury text on prototypes, which sometimes are not converted to text

Voted for it but I believe this feature should be introduced as a choice given upon scene import. Because there are also lots of cases when you use fonts that aren't pre-installed on users devices. So, keeping them as shapes is safer in these cases.