I would imagine the subtle daily increase or decrease in daylight, even in extreme latitudes, is more natural and healthier than an immediate one hour shift in either direction.

Hmm, instead of an hour shift, a 30-day shift of 2 minutes per day would be an interesting experiment. Hey why do I see so many raised pitchforks all of a sudden?

Getting anything figured out would be a mess, obviously, if everything were synced up (from the clock on your grandma's kitchen wall up to the clock on say Air Force One) it'd be easier (hmm, I see even more pitchforks). Someone's daily meeting with overseas colleagues would be e.g. 9:56 today and 9:42 in a week's time.

Another alternative would be 4x15 minutes jumps, every Sunday in a month.

Apparently you're not alone: https://gregcochard.com/daylight-savings-time-smearing/

Couldn't you hypothetically just do this for your own purposes? Write an app that moves your alarm time by two minutes a day and tells you when to go to bed?

You would have to choose a wakeup time that was conducive to your work schedule, even when you're getting up an hour "late" (right before the time change), but maybe it would still be worth it?

If we simply adjust the clock back half an hour every month we can eliminate leap years.