First post, I know but I found this little gem and was wondering if anyone had tried it on unsubstituted benzaldehydes?

Check this out!!
(Rated as: excellent)

To a three-necked 3000ml rb flask with a stir bar add 166g (1mol) 2,5-dimethoxybenzaldehyde, 15g Aliquat 336, 159g (1.3mol, 139.5ml) methyl 2-chloropropionate, 500ml toluene and 200ml MeOH .With a cold water bath the temperature is brought down to 10-15 deg C. 60g solid NaOH is added in portions of 2-4g over two hours while not allowing the temperature to rise over 15 deg C.

Do not get to brave while adding the hydroxide because the temperature can rise very quick, and you do not want to mop up the haloester from the floor and ceiling – trust me.
When about about 30g NaOH has been added the solution becomes very thick from some precipitate but this will gradually dissolve again. If it becomes too thick to be properly stirred add some more MeOH.

When all NaOH has been added allow the solution to stir for another two hours at room temp.
Then add a pre-made solution of 100g NaOH in 450ml MeOH dropwise during 2 hours while keeping the temperature at 20-30 deg C. Keep it stirring at room temp for another 4 hours and the let it stand over night. The next day the solution will be very thick from precipitated crystals. Add 500ml water and the crystals will dissolve. Separate the two layers and extract the organic layer with another 500ml water. The toluene layer can now be discarged.

To the combined aqueous extracts conc HCl is added until pH 2.5-3 is reached. This will cause a yellow oil to fall out. Allow the acid solution to stir for one hour at room temp, add solid NaCl (enough to saturate the solution) and 500ml toluene and stir for 10 minutes.
Separate the layers and extract the aqueous phase once more with 250ml toluene. Wash the combined toluene phases once with saturated NaHCO3 solution, dry with MgSO4 and remove the toluene by distillation. This leaves 2,5-dimethoxyphenylacetone as a yellow oil.

Yield: 152g (0.78mol) 2,5-dimethoxyphenylacetone

That's on Rhodium's page:

http://www.erowid.org/archive/rhodium/chemistry/25dmp2p.html

and this is what he said about it:

Rhodium
(Chief Bee)
07-17-02 20:24
No 333858
                  The base grabs the halogen from the alpha-halo ...        

The base grabs the halogen from the alpha-halo ester, forming a carbanion, which adds to the carbonyl, forming the epoxide. The reaction is a Darzen Condensation, and you should be able to look up the detailed mechanism either online or at the library.

The epoxy ester is then hydrolyzed and the formed acid decarboxylated. During the decarboxylation, the epoxide also rearranges to the ketone.

the articles that were retrieved for Barium that helped him to figure this out are still available possibly 

ok after sorting out the ref policy here are some articles :
Synthesis, 2008, p. 1221 - describes darzen condenstion of aldehydes and decarboxylation for a variety of benzaldehydes with methoxide, 2-chloropropionic acid and carbonyl with toluene. yield of unsubstituted benzaldehyde according to the authors is only 36%
JACS, 1947, p. 469 - describes a darzen condensation with o-tosyl esters.
Experimental for O-Tosyl Lactate -
To a solution of ethyl lactate (1 g, 8.47 mmol)  in 5 ml of dry  pyridine was added portionwise TsCl (1.77 g , 9.32 mmol) at 273 K. The resultant mixture was stirred at 273 K  for an hour and then at room temperature for 2 h.   After completion of the reaction, as indicated by TLC, the reaction mixture was diluted with EtOAc. The organic layer was washed with water, dried over anhydrous Na2SO4, and concentrated under reduced pressure to give the crude compound. Purification of the crude compound by column chromatography over silica gel (100 – 200 mesh) afforded pure ethyl-2-(tosyloxy) propanoate (1.05 g, 45 %). Physical data :1H NMR [400 MHz, CDCl3] d 7.78 (d, 2H), 7.30 (d, 2H), 4.10 (q, 2H), 2.46 (s, 3H), 1.50 (d, 3H),
1.21 (t, 3H).