It seems that some universities are OK with one other propose two and some even ask for teams. What is the standard? What is good practise? What would you recommend?
need - for what? To make the process work well? In case one dies? For references afterwards? I think that at least 2 is good, and 3 can be useful. A lot depends on how well the supervisors do their job and what their skill sets are. Ideally you want to have some different opinions and skill sets. A researcher with contacts in industry and in academia is in a stronger position than one with only one contact. Many institutes mandate at least 2. Some require a more senior academic as well - making 3.
I had two and they were absolutely marvellous. I was so lucky. They gave me continual support with my PhD and have left me with very happy memories.
They were very different both in experience and how they worked with me.
I spoke to quite a lot of universities before I decided on this one, but in my initial meeting with the more senior supervisor of the university that I chose, made me certain that it was her that I wanted to be supervised by. From my initial research proposal, she started to make sense of how I should proceed and there was no turning back. I knew that I was choosing just who I needed to be helped by.
I had only one supervisor while writing my Ph D dissertation. And, to be honest, on some occasions i would have dispensed with him. As the time went by, I found that he was an addictive man in some "hard" drugs. After all, I was successful!