Temperature affects chemical reactions, so the answer is a technical 'yes'. However, for you to actually notice it over all your tests, it would need to be a fairly strong effect over a small range of temperatures -- PD or the Enlighten crew would probably be able to give a good indication if this is something that will affect your results in general. I suspect that since reaction time is not usually used in identification of substances in a test sample, it will not change the specificity of your reagent test, just the speed at which the results are obtained.
Although I haven't specifically A/B'ed reactions at various temperatures, I tend to agree with BTs comments. Marquis reagent reacts via the following simplified mechanism (or similar condensations; see below), whether we're talking about reaction with amphetamine, MDMA or heroin.
As can be seen, heat is required to form the carbon-carbon bonds with formaldehyde. However, the initial heat evolved from the dehydration caused by the acid will usually be sufficient. It is likely that Marquis also forms C-O-C bonds with certain compounds so the effect of doing the reaction in a cold environment or with a cold reagent may result in a mixture of compounds reacting at different rates. This could obviously be helpful in recognising a suspected mixture of compounds. BUT, as this hasn't been confirmed in a laboratory setting where intermediates can be analysed and kinetics measured accurately, it should not be relied upon absolutely.
Mandelin is best kept at room temperatures to avoid possible precipitation. If the opportunity arises whereby it is possible to do these tests, results will of course be posted. In the meantime, as hoptis has stated, reaction speed should not be used as an indication of strength or purity.
as any of the enlighten crew who've tested at certain Melb winter events can confirm, the cold slows down a reaction hugely. but remember we don't put any emphasis on the time a reaction takes to occur, only the final result. well, within reason of course...
Despite the excellent scientific explaination offered above, I'll just quickly mention that I've anecdotally noticed a difference in reaction speed also. Given that the ambient temperature doesn't change to wildly here (in the UK), the difference needed doesn't really amount to a great deal... less than 30 Farenheit?
I was wondering, on a similar note, does the ambient humidity have any effect?
Without finding the references KostoN mentioned, I'd say this would be less so with Marquis as the added formalin already contains ~60% water. However, a moist sample may result in "clumping" which could effectively slow things down a bit. Mandelin contains little water so some reactions may be more affected by humidity.