3) immutability measurably does not reduce bugs in programs

I'd be curious to see how you back this claim up. Are you referring to something published that we can all go and read?