No, there's always a reservoir of variation; there never was a "clean slate." If it gets too small, the population is endangered--think cheetahs. For example, for diploids like us, recombination is a massive contributor to variation. Another source of variation is simply outbreeding; hybrid vigor is a real thing.
It's a fundamental misunderstanding of the mechanism of Darwinian evolution that makes it harder for laypeople to understand. It's always happening. Populations aren't waiting around for new mutations to happen. That's even the title of one of Behe's chapters!
That is indeed all Darwin was saying about variation. That's why it's misleading to plaster him with "random mutation." But Darwin pointed out that selection acts on that variation.
I don't see why. People are still looking for directed mutagenesis.
Do you realize that mutations are only random in a single respect--wrt fitness? That they are clearly nonrandom wrt location and direction?
Readily observable and directly measurable heritable variation. Otherwise, consistency would demand that you advocate for mutagenesis instead of outbreeding for endangered species.