We're neither latency- nor bandwidth-sensitive, and both services are overall quite reliable, so I haven't actually paid much attention to that aspect. But ther

We're neither latency- nor bandwidth-sensitive, and both services are overall quite reliable, so I haven't actually paid much attention to that aspect. But there are plenty of war stories of people using ZT or TS to do absolutely bonkers stuff, like zero-downtime migrations between cloud providers, with a write-heavy DB in there. I wholeheartedly recommend doing your own benchmarks! Everyone's workload will be slightly different, ultimately you should pick what works best for you.