zero system test fails often

The zero:check lookups against TTL=0 records system test fails fairly often on main and v9_18 in the CI (tho I was unable to reproduce it locally):

I:zero:check lookups against TTL=0 records (1)
I:zero:failed

Looking at the code there are two possibilities to fail the test code:

1. Either of six dig -f query.list instances fails to finish in 69 seconds (timeout=$(($(wc -l < query.list) / 5))), or
2. status: SERVFAIL is identified in dig stdout logs (dig.out$i.[1-6].test$n).

Looking at dig.out$i.[1-6].test$n logs I am certain that they don't contain status: SERVFAIL for any query from the query.list file. Perhaps one of six dig -f query.list instances is stuck, waiting for being collected by kill -TERM "$pid1" ... and wait "$pid1" || ret=1?