Do you have any reference to vanillin working? or any reason to think it does?

You must not confuse those with the color reagents, such as the Duquenois-Levine (vanillin/acetaldehyde) reagent used for detection of cannabinoids or Kellers, or Dragendorffs reagent which are usually employed for ergot alkaloids. What you are speaking of is visualization agents for TLC.

Method for the extraction of lysergol and ergot alkaloids from plants of the ipomoea genus
United States Patent 3920663

"by using as the solvent substance methylene chloride-methanol-benzene (25:5:5) and as the detector a 3% solution in alcohol of vanillin and 0.5% by volume of concentrated sulphuric acid, after activation at 110°C - 120°C during 5 minutes, exhibits the presence of several ergolic alkaloids, among which lysergol predominates, and, secondly, chanoclavine. "

http://www.freepatentsonline.com/3920663.html

The Chemistry of Peptide Ergot Alkaloids. Part 2. Analytical Methods for Determining Ergot Alkaloids
E. L. Komarova1 and O. N. Tolkachev1
Pharmaceutical Chemistry Journal

The other well known color reactions in concentrated sulfuric acid medium involve glyoxylic acid (blue-violet color), vanillin, Kellers reagent (purple-blue color), and Van Urks reagent (dark blue color). Most widely used are the condensation reactions with p-dimethylaminobenzaldehyde (p-DMB), employing Ehrlich's, Allports's, Van Urk's, and some other reagents. This method offers high sensitivity, whereby intensity of the coloration in the reaction solution is proportional to the content of the ergot alkaloids"

http://www.springerlink.com/content/w74k3xvxm68mm6g1/

For ergot alkaloids, it also looks like there is a modified Van Urk Reagent...



See here: http://aem.asm.org/cgi/reprint/45/4/1165.pdf

A colorimetric assay with 2
ml of Van Urk reagent as modified by Allport and
Cocking (2)

2. Allport, N. L., and T. T. Cocking. 1932. The colorimetric
assay of ergot. Q. J. Pharm. Pharmacol. 5:341-346.

Also:

Selection technique. Two types of selection criteria
were employed: the amount of alkaloid produced by
the mutant and nutritional deficiency.
Selection based on alkaloid production was accomplished
by utilizing the Salkowski reaction (11). In this
method, the mutant colony is replicated before the
experiment, and the reverse side of the original colony
is then scraped until no traces of agar are present.
After the reverse side is sprayed with a reagent
consisting of 1 ml of 1.5 M FeCl3 solution and 100 ml of
60%o sulfuric acid, the colony is incubated at 60°C for 5
min. The intensity of the blue color which develops
corresponds to the amount of alkaloid present.

The Salkowski reaction was used to differentiate
two groups of alkaloid production mutants.
High-alkaloid-producing mutants became dark
blue to black and, upon submerged culturing,
produced alkaloid in titers of >1,200 ,ug/ml of
culture filtrate. We isolated 28 of these mutants,
3 of which produced alkaloid in titers of >2,000
,ug/ml of culture filtrate. The 13 low-alkaloidproducing
mutants isolated were light blue to colorless in the Salkowski reaction and produced
<300 p,g of total alkaloids per ml of
culture filtrate.

11. Krebs, K. G., D. Heusser, and H. Wimmer. 1969. Spray
reagents, p. 875. In E. Stahl (ed.), Thin-layer chromatography,
2nd ed. Springer-Verlag, New York.

So the Salkowski reaction should be of interest --  If you can get any of those articles, please share them.