Edit. But I guess you were talking about our perception of time.
And, somewhere in our brain, we probably have just such a periodic event generator, although it's probably a bit flawed in the implementation like most things generated by evolution typically are, but it does the job well enough for us to arrive at a sense of time that very roughly agrees with more rigorously-derived measurement devices.
Ultimately time is a shorthand for cause and effect, and the fact that causes propagate to their effects at roughly the same speed (relative to a reference cause and effect) everywhere we measure them (relativistic effects notwithstanding). If you have evidence this is not the case, I would love to see it, because it implies some new physics. You would win a Nobel prize for this, easily, if you could prove it.
But you are right, if there was a outside observer the same stands, just for that observers observer -- hypothetically.
> In a world where physics depends upon time, is it even possible for time to not be steady?
I think so. Physics is a construct that we have to describe how our universe works.
Lets think about things in "ticks" where a "tick" is just one particle or I should say interaction happening in our universe. Or better think of our universe as a big state machine. Between each new state there are many ticks arranging the matter and getting ready for the next observed state. And by observed state I mean observed by the universe -- or realized. Now we have two time domains to think about. One, tick time. How long is a tick? Would it matter if some ticks took "longer" than others. Then you have the notion of time outside of a tick, that gives ticks its "tick time" or what makes some ticks "longer" than others. If this is anywhere near the case then time outside of the umbrella of each realized universe state does not matter.
I may not be doing a good job of getting the notion out of my head. But try to think of it if you were in a video game. Lets take it back to maybe the "Max Payne" days given it is a game that has some dealings with time. If you were playing the game on a laptop and closed the lid and the computer suspended would Max be able to detect or care even that the time between rendering one frame the other went from something like 60FPS (well back then 12FPS) to 2 hours our observed time to render the next frame? We would be the hypothetical observer. Max would not have a clue.
Anyways thinking about how time works is always fun. About all we can do is learn how time works from our observable position.
I can of course imagine a scenario where different entities, even in the same universe, experience the passage of time differently. For instance, a positronic robot whose consciousness is only switched on for one minute of every hour. To this being, events could appear to happen willy-nilly, without causation, due to activity between ticks, but the robot would have no way of knowing why the world works like this without talking to another being without the limitations who has been making observations in "real time". So perhaps something like this is happening to us and could explain some weird quantum phenomena, but I'm having trouble thinking of any way we could test it.
Yes, the problem of how to test such things. That is why it is so fun to think about it.
For all we know between the seconds a trillion trillion tiny life forms that resemble ants slip out between the cracks of space time and take a untold amount of time (to their reference) to nudge each particle just so according to a small tattoo on their forearms. When they are all done they slip away just as mysteriously as they came. And each time they come back to arrange our universe it is done so by ancestors of many generations removed of the last workers who came to enforce our its will on our universe.
This also brings me back to Lawrence Krauss? with the notion of there is not really nothing in nothing and we can't truly ever have a "vacuum" void of all. I think it was in his book Something out of Nothing.