Random safe-cracking

Dawkins posits thirty-nine rendezvous points for common ancestors (concestors) where the human line splits off from other major animal groups. Yet there is no evidence for these concestors. Not one has been found. Not one is alive, and not one fossil remnant exists of any of these supposed concestors. 

Chance is the rate-limiting step of evolution by natural selection. For example, consider a series of three doors in a hall, each door opening to the next, with a digital push-pad, combination lock next to each door. Once you have entered the proper code into the first lock,  natural selection can "fix" that code in place and hold the door open so a whole population can pass through and work on the second door. Clearly, the most time-consuming part here is the random sampling of combinations to find the correct code. Genetic mutations are like random sampling of combination codes.

All the evidence for evolution is for microevolution: rearrangement of what is already present in the genetic pool; no new complexity. Indeed, every example of evolution that Dawkins presents in The Greatest Show on Earth is evidence of microevolution: alteration without innovation. 

A group of Lego bricks can be rearranged, but can it create more Legos?…You might be able to make something "new" by removing Lego pieces (deleting/harming genetic materials–a frequent strategy in antibiotic resistance), but destroying does not help with the problem of creating. What if a friend gives you new Lego pieces. Well, this still does not solve the problem, because now we need to know where he got his pieces, and, in addition, studies show that natural selection tends to throw away duplicate pieces. 

In reality, we have oversimplified the whole situation for the purpose of clarity. Anyone playing with Legos knows that you can take the whole thing apart and put it back together however you wish. Biologically, though, there are rules about how the Lego pieces can be altered. For example, every single alteration or change has to result in a functional Lego model. Biologically speaking, functionless models are discarded. As you can imagine, this would severely limit the kinds of variation that could occur. There are many ways to arrange five Lego pieces, but only a handful of configurations make something useful. The useless ones are not preserved and cannot be used as intermediates on the way to building a useful model. This is sometimes known as the "sampling" problem. 

Even the first step of generating an innovative protein fold in an existing protein (something which must have been done at least 1,200 times in history) seems increasingly unlikely, if not impossible. 

Changing gene expression merely means changing the amounts of already existing proteins, but a new ability sounds like innovation. However, genetic analysis revealed that the bacteria already had the ability to eat citrate, but it was normally suppressed. So, in the end, "No new genetic information (novel gene function) evolved". Furthermore, many of the changes observed in Lenski's bacteria result from loss-of-function or decrease-in-function mutations, which is the exact opposite of complexity creation. And so, after 45,000 generations, still nothing new. R. Poythress, Richard Dawkins (P&R 2018), chap. 5.