Wow, that is nice that such easy routes of reduction are available. Btw, you may also use thiol reduction with Ni raney (without hydrogen). It is known for thioacetals, but something similar should work with sulfides. The procedure for merkaptals is as simple as following:
"The reaction of hydrohenolysis is being carried out by refluxing thioacetal in 70% alcohol with Ni raney for 5 hours"
You can make sulfide from thiol and haloalkane, but then there is a side reaction of aziridine formation possible(i should inquire, not very positive). It is quite possible, that thiocyanate + 2-aminopropane-3-ol followed by HCl reflux yould yield thiazolidinone, that after hydrolysis(or without it?) can be reduced by Ni raney. Still not sure about thiazolidinone formation, besides benzylic OH would react faster, you neen to get rid of it wirst or also protect it

Headstrong, please look the article below, it may be of your interest, as it deals with reactions of 1-phenyl-2-amino-1,3-dihydroxypropane and production of 3-amino-2-phenylthietane which is one step converted to amphetamine(Ni raney/ethanol, without H2). Specially, ctrl+F for words "Dissolving the cyanothiamine", look pages 32, 29. It seems you may avoid using HI, P and reduction of terminal I, and stick with thiourea or alkali metal thiocyanate and base followed by Ni raney. And it seemed from the article for me that you may not even need to use HCl, just diol and thiourea or thiocyanate , but here i may be wrong(lost the page where i probably seen it) and you should first convert benzylic position to Cl and then treat with thiocyanate. Anyway, it is much easier then HI route. Also google for 3-amino-2-phenylthietane, there are many refferences of their synthesis as they are mao inhibitors
https://circle.ubc.ca/bitstream/handle/2429/20317/UBC_1977_A6_7%20K35.pdf?sequence=1
(the pdf is 6mb, better use right button and save as)

You always respond almost immediately! this article is just for information, that such method exist. it does not need immediate respond as it is big and you can't read it immediately so, feel free to respond or not respond any time you like.
Function of thiol is that it forms thioacetal with aldehyde R'CH(SR)2 that is further reduced. i dont know the exact mechanism of Ni raney desulfurization(you can google it) but it is seemingly radical. but instead of thioacetal you need sulfide made from thiol or H2S and haloalkane. I recognized the method that avoids using chloroalkane, it is this one http://pubs.acs.org/doi/abs/10.1021/jo00832a032 but making cyclic carbonate from OTC reagents is a problem, unless you have CCl4 available in some glues, or if it can be done with urea the same way it reacts with propylene glycol to yield propylene carbonate. Both reactions may not work, also because of nitrogen between 2 OHs. Or there is a slight possibility that oxazolidinone, if it would form instead of cyclic carbonate in either of routes, especially the product of esterification of benzylic oxygen, can be opened by same nucleophilic attack with thiocyanate, and then it would form thietane as on scheme.
I understand your point about chemicals. But if the only source of sulfur in your area is thiosulfate, no thiocyanate, then this route would be harder(using KCN). Other routes for making thietane are also not very kitchen friendly, you would need 2 times chlorinate alcohols with HCl if you want to use thiosulfate(first benzylic OH to Cl, then thisulfate treatment, then primary OH to Cl, then cyclization with base). For reducing only primary OH, you need any thiol and primary alkylhalogenide that you would make using HI. And then Ni raney. If there is no possibility to find Ni raney as well - better to give up this route with sulfur reagents, because you found quite convenient one using metal reduction of terminal Hal and there should be no problems with matches for P and pharmacy stores for I2. Or, try to find the most convenient route, write out all not available chemicals in your area and travel those 300 km and buy many things that you want to experiment with.. i don't see other options, as number of syntheses from only OTC reagents is limited

Concerning thiocyanate, it seems it can be done OTC as well. It forms while prolonged heating of thiourea http://pubs.rsc.org/en/Content/ArticleLanding/1974/P2/p29740000896
and thiourea can be made from hydrogen sulfide and cyanamide(can be made from urea and CaO http://www.sciencemadness.org/talk/viewthread.php?tid=14267). Hopefully in thiourea synthesis hydrogen sulfide can be changed to alkali metal sulfide, but i'm not sure. Hydrogen sulfide is toxic, but can be made by several routes OTC.

And IMO it heating KOCN and excess of Sulfur will result KSCN as well, KOCN can be easily made by heating urea and KOH. Kitchen friendly

it would be a nice reaction if it really works, sulfur melts easily so the surface contact with oxidizer and reductant is substantial, but it evaporates at about 400, and that is not that good, loo low temperature for reaction i suppose, but maybe it is sufficient or some special catalytsts should be used. i failed to find anything in google but it does not mean such reaction is impossible

As for diiodo route, i think it is possible then, and should be the easiest one.
But the clemmensen route on this substrate may give many by product, because those two carbonyl groups would be present as conjugated system (one CO is in a form of enol and its double bond conjugated to second carbonyl). Such conjugated systems, i suppose, would react not normally, and even if they do(maybe there is osmething in literature) - aminogroup can as well become a ketone instead of terminal OH and split off, so you can get a mixture of 3-phenylpropanol, 1-phenylpropanol and 2-phenylpropanol along with amphetamine. Amine can be easily separated by acid-base extraction but its yield may be not high also. This is a proposal, but if there are examples in literature then it is ok.

why don't you try to reproduce procedure in reply #11 in this thread, it seems 0.5h is sufficient

Ho hum...I just looked at the structure of chloramphenicol, and hey, a hoffman degredation of the acetamide group through its intermediate isocyanate in aqueous medium, protection of the resulting N-methyl amine followed reduction of the nitro to an amine, then deamination of the aryl amine group with HNO2 in situ from NaNO2 and dilute H2SO4, deprotection then reduction of the two remaining hydroxyl groups on the sidechain with the classical hydriodic acid route of choice in 2x molar amount, and what do you get? damned near OTC amphetamine.  

wat

Start from phenylalanine instead; you'll have to perform an extra reduction, but you'll save two reductions and a hydrolysis.