The precursor experiment produced an interesting data set:

The data can be subdivided into positional effects and chamber effects. The complete data set indicates a trend that shows better when the data were grouped and summarized properly.
Adding all data per indiviual chamber clearly showed that the best germination resulted when a volume of 500 µl was dispened into each seed casette. The type of paper had no effect at the highest volume. However, comparing normal paper (N) with heavy paper (H) and a combination of both (N+H) showed that two layers of paper reduced germination at lower water quantities.
Similar to the previous figure, combining the paper data regardless of water volume, also showed that one layer of heavy paper is better for germination than two layers of paper (N+H).