Number of seconds is not a relevant metric for some people.
More relevant for me is how much and often I need to switch my physical and cognitive task at hand to briefly use my device. If I'm in a focused mode going back and forth in my house getting things done, it's a huge benefit to just verbalize my thinking in the moment rather than come to an abrupt stop and switch to visuospatial navigation to drill down to the timer app, and fine tune the amount before I can switch back to what I was doing before.
But that argument only holds if the voice assistant actually works. The point is that it often doesn’t, requiring repeated attempts that end up taking more focus and time then just manually doing it. There’s no way that doesn’t end up requiring as much if not more cognitive switching.
More relevant for me is how much and often I need to switch my physical and cognitive task at hand to briefly use my device. If I'm in a focused mode going back and forth in my house getting things done, it's a huge benefit to just verbalize my thinking in the moment rather than come to an abrupt stop and switch to visuospatial navigation to drill down to the timer app, and fine tune the amount before I can switch back to what I was doing before.