Avoid following logs for Tendermint in CI

which does not work with multiple instances.
Instead debug start logs.

scripts/tendermint/start.sh

@@ -46,6 +46,7 @@ if [ -n "${CI:-}" ]; then
   # Give process some time to come alive. No idea why this helps. Needed for CI.
   sleep 0.5
 
-  # Follow the logs in CI's background job
-  tail -f "$LOGFILE"
+  # Debug start
+  sleep 3
+  cat "$LOGFILE"
 fi