The name @ corresponds to the empty subdomain (= it is for records which apply to the domain name itself). You can simply leave the subname field empty.
I am unable to reproduce the problem from several other vantage points, and all our frontend nodes are up and running, and I succeeded in connecting each of them individually. As a result, it appears that there is a routing issue between these services and ns1.desec.io.
To debug the routing problem, I contacted the Zonemaster admin and asked for some routing logs as seen from their perspective. I’ll keep you posted on the progress.
There was a routing issue in the US that caused these problems. As the services you were using seem US-based, those were affected by the problem.
Sorry it took quite a while to fix! Meanwhile, ns1.desec.io was available from other regions of the world, and ns2.desec.org was available globally throughout. Apart from a bit of extra latency seen in certain US locations, there was no other service degradation or disruption.