3-methoxymethamphetamine is a monoamine releaser that is not explicitly scheduled; however, it is a positional isomer of an internationally scheduled substance (4-methoxymethamphetamine). Some reports suggest that it is subjectively similar to MDMA. Phenylephrine is decongestant that has replaced pseudoephedrine in some products.

Here is my scheme: I've only included two references (both for the dehydrogenation), because the other reactions are well documented. That oxidation/dehydrogenation is the sticking point. Can anyone find better refs?

3-methoxymethamphetamine is a monoamine releaser that is not explicitly scheduled; however, it is a positional isomer of an internationally scheduled substance (4-methoxymethamphetamine). Some reports suggest that it is subjectively similar to MDMA. Phenylephrine is decongestant that has replaced pseudoephedrine in some products.

Here is my scheme: I've only included two references (both for the dehydrogenation), because the other reactions are well documented. That oxidation/dehydrogenation is the sticking point. Can anyone find better refs?

I suggested a route like this over at wetdreams.
The problem is, that phenylephrine is actually present in really tiny amounts in decongestants. Even from chemical suppliers it usually costs around $10/g, which is hard to justify considering the rather low potency of the final product.

The one other thing you might want to consider is that the formed imine is hydrolytically unstable, and so you ought to run the Grignard immediately after generating the imine, since the Grignard requires anhydrous conditions anyway.

This implies that you'll want to methylate the phenol before forming the imine, or you'll have to work with the phenolate salt to form the Grignard, which is going to be problematic due to low solubility in ether. Methylating a phenolate in presence of an amine requires a pretty selective reagent, though instinct says that quaternary ammonium compounds might work:
http://dx.doi.org/10.1016/j.tet.2008.10.024

1-phenyl-2-aminopropanols are known to somehow undergo the similar rearrangement as corresponding diols do. i'll try to find a ref, or a place where i've seen it. this may save 2 steps.

Now this is interesting!

maybe phenolate would be methylated faster then imine, but imine is also able to be methylated, maybe not in that extent(i mean, imino function is not used as a protection group to avoid methylation, afaik. nitrogen does not lose nucleophility).
about my above post, is something unclear? i meat practically same you said, that oxidation to imine is possible but there are no mentions of aliphatic imines formation in the refs you'd given, so some other methods or refferences are required.
What about 3-MeO-MCAT? is there any information about it? If OH group is first oxidized to C=O, then a-position of this ketone would be enolizable and possible to methylate, provided nitrogen is protected and the corresponding refferences found. If it is possible, then phenolic OH group and a-position can be methylated in one step. Nitrogen can be protected before oxidation, this would protect it from methylation as well. I wonder if a copper chelating of aminoketone would make protons in a-position acidic enough to be quantitatively abstracted by alkali and the resulting anion to be a nucleophilic enough to react with MeI. Otherwise, a nitrogen protection and a strong base deprotonation should be used