Having 1 track and 30 VSTs results in high CPU usage - heavy load on one core. Distributing over 8 tracks is highly recommended as you are balancing across all cores. The latter method is much more stable at low latency. 1 track with 30 VSTs presents higher risk of audio engine "crash" - crackling and audio dropouts.
This is a discussion of serial VS parallel processing. If you have vocals on track 1, guitar on track 2, drums on track 3 etc. and they blend together at the Master / Reverb bus - the sounds are distinct and clear. You have better control over each inputs...... Serial processing where everything is on 1 track: VST 1 feeds into VST 2 feeds into VST 3 carries the risk of distortion. When one part is distorted, everything downstream becomes distorted.