Do you not think that solutions to erdos problems might end up stepping stones to other important problems?

Either by introducing new tools, or by proving things that were previously unproven that end up helping in unexpected ways?

That's often how math goes, isn't it?

This is, indeed, how math often goes.