Trying to read the paper... I guess if you ignore the difference between finite and infinite tape Turing machine, and if all physical constraints are outside the scope of the paper, then it is easy to prove the universe can simulate itself.