In general I am a big fan of processes, not big bureaucratic processes, but light-weight processes like those in the scrum framework. A good process provides visibility into ongoing efforts, and some degree of certainty in the outcomes.
Processes are not "one size fits all" however, and are not applicable to every situation.
Processes make sense when the inputs into them are predictable, and the desired outcomes are clearly defined, namely everyone agrees what success looks like.
Conversely, when inputs are chaotic, and the desired outcomes are uncertain, a process is going to add very little value at all. It may even be a negative to apply an existing process to an unknown situation, as it can lead the team into a false sense of security about the outcomes.
Put simply, processes do not protect us from chaos.
In an uncertain situation, processes need to be invented from scratch, for example before anyone flew to the Moon, there were no existing processes on how to do that. They had to be invented.
For leaders, it is very important to spot when an existing process will work, versus a whole new process being required. It is also important to think about who is applied to the problem: junior folks can execute an existing process, but senior folks will definitely be required to navigate through uncertainty using their best judgement and experience.