I do my "long" runs on Saturday. It's just how my schedule goes. A long run for me is 10 miles or longer. January in 2021 has five Saturdays, so I had five long runs. As you might imagine, running that distance takes a while, and if you're a data-driven, fitness tracker wearing nerd like myself, you know you score big points on these kinds of days. The kind of numbers that will screw up the average. More on that later, but first here's the first long run on January 2nd.
The idea of a daily average is probably fine for most people, but I would like a little more control. My average Tuesday is nowhere close to my average Saturday. I suspect the goal setting system in Apple Fitness, and probably many other competing products, use a daily average to create challenges. People on a training plan will likely have wildly different daily needs because in any plan there are "rest" days. It's not that you do nothing on these days, just that you do far less so that your body has time to recover and repair the muscle damage from your workouts.
These fitness systems we're using have all the data they need. It doesn't take a massive amount of machine learning. Apple Fitness even shows daily averages in the app. Right now it's tucked away in the Trends section. Who knows where it will be by the time you might be reading this.

So why can't our devices give us better daily goals and instead treat every day the same? Again, maybe that's more than most people need, but it wouldn't it be a nice option with the added benefit (to the system maker) of showing how "smart" their systems can be?
I'm most familiar with Apple Fitness and my device is an Apple Watch (Series 5 for the moment). The system has evolved nicely over the years as they discovered the market for Apple Watch was primarily fitness. I'm hoping what I'm looking for is on their "roadmap" (a term I'm really learning to despise in the context of software development).
But what about that whole five Saturdays in January bit at the start? When you're looking at month over month (MoM) or Year over Year (YoY) comparison, if you're not taking into account that the composition of days in a month change from year to year, or that both the composition of days and number of days in a month change, your projections can suffer. January 2022 will still have five Saturdays and I'll likely do five long runs, but what about when there are only four Saturdays? That will a sizable impact. By then, I hope my fitness overlords have figured out how to generate realistic goals for me.