The issue is that even if it's very much milder than Delta and only a tiny % become very ill, if it's as contagious as it looks, a LOT of people will get it. And a tiny percentage of a lot of people can still fill up all of our intensive care beds, not to mention take a large % of people out of work for 10 days or more
The NHS has very low margins of beds available for new admissions - usually about 5%. There are currently about 5% of occupied beds with covid patients. It won't take much of an increase to push NHS into a real crisis, especially if lengthy admissions are required.
The small percentage of a large amount is a number that is concerning to planners. Even if omicron causes 50% fewer admissions, that only gets us the grace period of one doubling (as current hospitalisation rate is about 2%, down from 9% before widespread vaccination)
As it appears to be doubling every 3 days, it could get very dicey very quickly.
We'll know more in a few weeks, as it'll take (based on disease patterns to date) a week for the new omicron infections (acquired this week) to need admission, and then a further week or so for deaths (though of course there are lots of cases outside those timelines - so it could mean at 3weeks before there is much chance of an adequate picture.