You'd have to figure out where the risk inflection point is on the chart for "spending 7 hours with 15 kids vs. spending 1 hour with 5 sets of 15 kids," but it's probably much less risky to be with that small group rather than cross-contaminating all day. The difference between being with an infected person for 1 hour vs 7 is probably pretty small, and if you're seeing multiple classes you're increasing your odds of coming in contact with an infected person.
The FDA gave EUA for test pooling up to 4 people. The Harvard doc on a recent episode of TWiV argued strongly in favor of test pooling. The idea is you can increase the throughput of the testing greatly by running 4 samples at once. In places with a low enough positivity rate, like Illinois, a big majority of the batches will come back negative, and now you've just reduced the testing time for those 4 people by 75%. If the batch tests positive, then you re-run the individual samples to narrow it down. It can be very efficient at screening down to the trouble spots. Not sure how well it'd work when you're in an area with 30% positivity rate like some states though.
https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-issues-first-emergency-authorization-sample-pooling-diagnostic