Loris,
I absolutely agree with you. I think I need to clarify a few points about the glass experiment too, there were also some other surprising consequences.
I have only done limited testing with commercial papers; I gave up on them when I realised that in one of the pads of water colour paper I had, every 3rd sheet behaved differently to the others and must have been from a different production run. I have not tried Fabriano 5, but I will as it sounds really interesting. I need to print some stereo pairs of images where my hand-made paper has too many surface artefacts and spoils the effect; I think the hot pressed variant could be ideal.
I do not have problems with the blue washing out at all and the cyanotype goes very dark, however I have seen other examples that are darker so you are right that my paper is not giving the Dmax possible from the process; the blues I have though are deeper than any commercial paper I have to hand. I have a few images where a real near-black tone would be very useful, but for many images I actually print for a slightly shorter time to rein back the contrast a touch.
I have attached an example scan I have to hand of a 10 minute exposure that has a nice dark tone, but not as dark as possible; I was comparing it to paper soaked in alum which reacts about a stop faster. For my maximum depth I need to give the print 20 minutes under my UV lamp (after at least a 5 min warm up). I have trouble scanning some of the images as when viewed in the hand, they have a metallic 'shimmer' that gives uneven tones in the scan (making producing photoshop curves a pain!)
I have only tried hardening the gelatin in alum as formalin is not so easy to get hold of here. The change is dramatic and I do start to experience loss of blue and also difficulty in washing, although the exposure time decreases. I can quite imagine the issues you have with the sizing for use with gum.
The effects of previous layers of size and also any additives in the paper do have an impact on later layers. The changes can be quite subtle though which makes comparisons tricky. One example was where I air dried and hair-dried half/half a sheet of paper. I exposed both halves at the same time under the warmed-up UV lamp and the printed-out images before development looked different. I washed them together in the same water and they washed at different rates too. After 24 hours when they were dry and pretty well oxidised, I could not tell them apart. If I had printed each one in the sun and gauged the exposure based on the print-out image, the final results of the test would have shown a marked difference.
The glass experiment was tricky as I normally wash in cold running water to prevent the paper base making the wash water acidic or alkaline (makes a big difference with pH12 paper as the image bleaches out if you tray wash). In running water, the gelatine slipped off the glass and went down the sink

Instead I immersed the glass gently into a bowl full of water in an attempt to dilute-out any washing effects. The paper I washed in running water. The first attempt that worked for glass had a fascinating twist in that the gelatine in the highlights still floated off, but the gelatine in the shadow areas was hardened by the cyanotype process. My hypothesis was that it was the UV light curing the gelatin in the shadows so I coated some glass and gave it an hour of UV before coating with the cyanotype solution. The gelatine in the highlight areas washed off just as before! I can only assume there is a similar interaction on paper but have yet to test it (I plan on the tinting the final layer of gelatine size).
My main conclusion is that experimenting with cyanotypes is just really tough! I have had to go to extremes to make many of the experiments repeatable from scratch and any change of variables can result in a very unexpected output. I still have not decided on a definitive way for making my paper and the corresponding curve if I am using digital negatives.
Best regards,
Evan