Wow I never knew Jenkins could scale its workers like that. We definitely have performance issues with our Concourse cluster running in K8s. It's got it's own set of dedicated nodes but we need to scale the workers better as they're ofter under heavy load throughout the day when the devs are pushing code and a bunch of tests are running in Concourse (and PR checks as well)!

Concourse definitely feels more refined than Jenkins. Like other commenters have said, it's a steep learning curve to grasp how things move between Tasks/Jobs.