I find little to argue with the Dave Woldrich Critique. Each point he makes could be a thread in its own right. I'd be pleased to participate in those threads. But first I will suggest a couple of reasons myself for wanting a redo from first principles.
We write programs differently now. Our programs sprawl over dozens or thousands of computers.
We write in dozens of languages at once. Each with its own special idea of what it even means to compute.
We don't get right or wrong answers, just good enough answers, sometimes delayed or drawn from a cache.
One thing hasn't changed: programs worth running have been methodically tested.