You do you, but here's why I don't follow this practice. I start my explanation with an old joke:
One evening, I spotted a man staring intently at the ground under a lamp post. "Looking for something?" I asked. "My watch," he said, and continued staring at the pavement.
"You lost it right here?" I asked, just to confirm. "No, I think I lost it in the alleyway back there" he replied. "Then why," I asked, "Are you looking for it over here?"
"The light's better."
This is the general problem of GitHub as a source of data for whether someone is a productive team player. If they did a lot of productive team playing in public on GitHub, great. But many people do their productive team playing off GitHub, or in private repos. They're like the fellow who lost his watch in the alleyway.
But we can only see what's under the light, and evaluating them as a complete developer based on what we find or don't find under the light is like looking for our watch under the light. We're doing it because the signal is easy to see. But that doesn't mean it's all of the signal, or the most relevant signal.
If you have a statistically significant data set showing that people with inconsequential GitHub contributions in public are inconsequential contributors in a professional software development environment in private, I'd like to see it.
But lacking a rigorous study establishing a statistically significant correlation, people's public GitHub profiles are not a reliable source of signal for hiring professional developers.