And btw, I agree with everything in the video. I don't know how you can find statements #2 and #3 above "unjustified"; the whole video was justification for those conclusions.
My 2nd criticism of 2 was that "We can invest in things we can never know. But to do that is nonsensical" makes a normative statement about a descriptive option which would require a normative system to make that judgement. Morality is the subject that best fits the category "things we can never know". Hence it invests in something it cannot know to refute investing in something it cannot know. (This criticism assumes the author cannot know the answer to the important question "What ought one do?")
There will always be certain moral dilemmas without a clear right or wrong solution, but I think there is enough that virtually everyone can agree on that I would not group the whole subject of morality into "things we can never know." That murder is wrong is about as clear as the sky being blue, for example. Making Pascal's wager also seems pretty clearly absurd to me -- not based on any system of morality but based on rationality.
All normative objects have 2 definitions:
A normative definition describing the moral character and a descriptive definition describing what the object is.
Murder is commonly defined by a universally accepted normative definition: "Wrongful killing."
However Murder does not have a universally accepted descriptive definition. Usually it revolves around unresolved questions like: "What can be wrongfully killed?" "What factors can justify killing?" ...
Murder, the most important aspect (from past humans' points of view) of the important question, has not yet be rescued from "things we do not know" why would you claim that morality as a whole is not still inside that category?
