The point is that that saturated field has absorbed a lot of water already. When it is full it is full and yes, when it is at capacity that will result in the water flowing further down the river system. But it takes a lot of water to get to that stage. If it wasn’t there, if you concreted it over, that water would also just gone straight into the river system rather than been stored in the field- you’d have had a bigger flood, and for longer.
OK, real world example.
We are working with a developer starts on a large site pre Covid, for 2019 to 2022 every year without fail would complain that their driveway flooded (unsurprisingly it's almost the lowest point other than the ditch next to it for some distance).
That site is majority built now, nothing for the last 3 winters, even when we were getting calls about were our systems suitable as the developer was concerned about the water level in some of their systems (turns out that was due to neighbouring farmland sending water towards the site due to a failed pipe which has now been rectified).
Soil has a bad capacity for storing water, yes there's a lot of it, but let's say it's around 100,000 litres per acre. However, that's a much scarier number than it needs to be.
1m3 of water is 1,000 litres, so a pond with a base roughly 4.5m x 18m (by the time you cater for side slopes that's circa 12m x 26m) will easily hold that water.
Now yes, that pond will empty out between rainfall (even if that's a bit) however that's why it works, because there's then more storage (maybe not all but some) for the next rain storm.
It's leaving that site slowly. Typically the goal is that (say going into chalk via soakaways) is that half that water would soak in over a 24 hour window.
If your sending to a water course it is likely to be a bit faster, maybe around 6 to 8 hours for the same output.
Ah, see, some might say, that's going to cause flooding.
The thing is that's 6 to 8 hours after the storm has ended and is 1/2 the water, with the rest maybe taking double that again.
Even fairly normal fields over 10% of the water which falls on it will run off.
That's not even too bad for a 1 in 1 year storm, however, make that a 1 in 10 year storm and you've got to deal with about 150% of the water. Make that 1 in 100 and it's over 250%, maybe as much as 350% of the rainfall.
Yet the drainage systems have to not only cope with 1 in 100 plus 40% and still let the water out at the normal rate, even though it's storing 350% to 490% of the water of a 1 in 1 year storm.
That's before the latest guidance which is requiring more rainwater harvesting (used to be that a rain butt per house was enough, that's not nearly enough now) and whilst there's always the potential that no water was used since the last rainfall, for summer storms (which can be nasty) it's more likely that it'll provide some more storage.
The real issue isn't the volume of water, it's how fast it gets into the place it's going to cause flooding.
If it all gets there in under an hour, that's when you have issues, when you stop 1% of it from getting there until at least 2 hours after everything else then the risk is reduced. Do that enough times and flooding can stop being an issue.