SilentCal comments on Tackling the subagent problem: preliminary analysis