It seems like the Russians have managed to circumvent the problem mentioned in the article above - that the propylene oxide is hydrolyzed before the alkylation takes place - by blowing dry helium gas through the reaction mixture to carry away any formed HCl.
Intact propylene oxide complexed with AlCl3 is alkylated by benzene in the least hindered position, yielding 1-phenyl-2-propanol.
If the propylene oxide first undergoes acid-catalyzed ring-opening, an isopropylic carbocation (carbonium ion) forms, which then attacks the benzene to form 2-phenyl-1-propanol.
So - by eliminating protic acids from the reaction mixture, the epoxide stays intact, and the more interesting of the isomeric alcohols can form. This is probably the explanation for the patent mentioning that "During this reaction HCl gas evolves, leading to low yields and an impure product, therefore He gas is allowed to flow through the reaction mixture."