Amigdalin is a glycoside, that is to say a complex molecule composed composed of two parts: a sugar (usually d-glucose) and an aglycon. The aglycon in the case of almond, apricot and laurus cerasus oils, is constituted by two molecules: benzaldehyde and cyanidric acid. Amigdalin specifically has this sequence: two molecules of glucose condensed in line, the whole condensed with a unit of benzaldehyde and cyanide by one side. This is the shape of the complex molecule. The bonds are not strong and can be broken easily: with an acid catalysis, I think, or by enzymatic cleavage. A glycoside is an trick of Nature to convey a toxic substance in a host organism. Many animal poisons are glycosides. Also the cardioactive toxins of the digitalis species are glycosides.
Anyway if it's benzaldehyde you're searching for, have you ever tried to buy synthetic bitter almond essence? It's supposed to be pure benzaldehyde in a solvent like IPA or a similar alcohol. If you search on the Net for catering suppliers you can find big liter bottles of the stuff. So you haven't the isolation problem. If you make the condensation with nitroethane to get the corresponding nitro-styrene you need it lab grade pure, but if you do the L-PAC biosynthesys, I think you can use it straight from the bottle, because the pyruvic acid in the acyloin condensation "selects" only the aldehyde molecules and the presence of the solvent is not a problem. At least I think so.