How "Real" is Your Real-Time Simultaneous Speech-to-Text Translation System?