It does not sound like a wide margin to me. 3,500M or 4,000M are kind of close targets, when you consider that the Earth has been a continually changing system

It does not sound like a wide margin to me. 3,500M or 4,000M are kind of close targets, when you consider that the Earth has been a continually changing system (unlike a rock such as the moon). Pretty hard to construct a timeline accounting for all the changes that might have lead any one of your supporting conclusions down the wrong path. When you revise one of your conclusions, you have to revise all the conclusions it supported, and it might turn out that one of your numeric estimates was 500M off. Easy. Astronomy by contrast is much more regular, so then a 500M year difference can indeed be a big deal.