But when we actually did it, it was basically fine. Obviously everyone still learnt the standard place value algorithms (by that point we already knew them, and they didn't stop teaching them to younger kids), but people also learnt other strategies like: if I'm adding 47 to 54, I can simplify that by instead adding 50 (47 + 3) to 57 (54 + 3), which is 107. It was about learning what you to terms like 47 + 54 without changing the value.
Another one I think I remember being mentioned was thinking of 124 x 52 as 100 x 52 + 20 x 52 + 4 x 52. Of course, that's exactly what you're doing with the standard algorithm! The 'algorithm' you learn when you're 5 is just the same as the 'strategy' you learn when you're 10, but presented as something you do on paper rather than something you do in your head.
I also remember learning about ways to approximate these sorts of things. So you want to work out 4241 x 1245 - well you can work it out, but you should probably also work out what 4000 x 1000 is and use that to check that your answer is of the right order of magnitude. And if you're just working something out quickly, then maybe all you need is 4200 x 1300, that's obviously close enough for any practical usage.
I don't remember exactly what I learnt then which to me is a sign it was probably very effective: I internalised the strategies, some of which I was probably already using, some of which I probably didn't remember but thought about again later. It all seems very obvious now, of course, but it's different when you're 10.