The most common form of petroleum ether consist mainly of heptane (with some pentane, hexane etc.). I'm looking forward to hear of yuor results. Try to stick to the published procedure as much as possible with no shortcuts whatsoever to make sure that in case of failure, it is not because of any unnecessarily introduced variables, making your testing of the method as credibly as possible. A failure by someone who doesn't even have a scale does not establish that the method isn't useful for ephedrine. Also make sure that your ephedrine is absolutely pure and doses not contain any other things, such as binders from ephed pills.
On what scale are you planning to perform the synthesis, and how do you plan to isolate and purify the product (as that is not included in the TSII writeup)? I would suggest that you scale the procedure by 5-10x the amounts in TSII. What kind of lab and analysis equipment do you have access to?
Do you have access to the original article (Tet Lett 28, 3817 (1987) I think) or just the experimental part from TSII? I would be interested in the exact reaction mechanism.