If the whole thing is such a big, crumbling Lego tower, and it's going to take three years just to fix the problems it has now, this doesn't seem like a compell

If the whole thing is such a big, crumbling Lego tower, and it's going to take three years just to fix the problems it has now, this doesn't seem like a compelling argument to keep using it. Surely at some point it must be easier to abandon it and start over - obviously that won't just solve all problems, but there must be a line where the cost of maintaining the current thing outweighs the cost of a rewrite.