If it is too risky to have students attend school in person, it seems to me that it is too risky to have any school-related activity be held.
I disagree on some key points here.
First, it is not really a binary choice between "safe" and "too risky." The choice is between greatly increased risk of spread in order to get some level of improved effectiveness of in-person classes. Since we are talking about high school activities, I think we have to focus on HS school students. I'd say most HS students are going to learn better in-person than online. However, outside a very small few that difference is going to be relatively small. If your HS student gets nothing out of online classes, chances are they are getting nothing out of in-person classes. I'm sure there are exceptions, but for a large majority this is probably the case.
Second, people argue as if in-person learning is incredibly valuable in public schools, and activities have marginal value to students. I think this is incredibly wrong. I would much rather have a HS kid learn algebra marginally less effectively and get the valuable lessons that sports and activities teach them in terms of dealing with adversity, working hard, working as a team, etc.
There is a huge difference between getting dozens of kids together for an organized activity vs. getting hundreds or even thousands together in a building. There is at least an option to take classes remotely. Most activities an alternative to in-person simply doesn't exist.
So I see organized activities as having relatively low risk with high reward. I see a large school putting kids in the building in order to get some level of improved learning as a huge risk for a marginal reward. So I see the peak of the reward vs. risk curve for high schoolers being 0-2 days in person with activities still going.