Hi DeuxVis.
After browsing the web, I found a few information about delamination. One of them was the advice to have slices no more than 60% of you nozzle diameter. Until now I always thought 80% was fine (and for PLA and on my Mendel with 110ºC heated bed it has been so far). Until now I had been working with 0,2mm heights, with my Reprappro 0,3 mm nozzle - BAD. So I changed my slicing height to 0,15mm.
Conversely, I found that delamination can also come from too fast cooling, so I disassembled my passive cooling 'guide' (switched off the minimum fan speed in Slic3r a while ago, as nicely explained in the instructions).
Now my current configuration is 80ºC heated bed (well, my IR thermometer tells me 72ºC to 77º in the corners of the heated bed) with hair spray (no need to respray for each print, I don't know yet how many prints you can make before having to prepare the bed again - though the stuff mightily stinks when you heat it the first time, on following prints you don't smell it anymore, although your parts have a whiff...) and 6mm brim (yes I have been working on this one for quite some time now: it is a lot, but seems to work pretty good). So with this I printed the z-top_left.stl successfully, with just two small slits (I won't call them cracks, although I have no doubt the part will break if I use it as is). Hopefully a friend of mine should be able to make pictures with his professional gear next week and I'll show the difference. Also, I will do some acetone-vapor curing and am curious to see the result (hopefully with pictures as well).
Right now I am printing a big cogwheel with V-shaped teeth, layer 61/93, so far so good.
For the point about strain not being the same everywhere: A+, this explains quite a lot. This is why some parts print better than others.
PS.: Peltier will definitely NOT work at so high a temperature, their internal temperature will be much higher and the inner solder will melt, destroying the Peltier. Right now I am working with 3 cemented 4R7 resistors. I just received my four TO220 7R5, I will have to test this by fitting them into the corners. Should be done next week as well, maybe with a new heated bed isolation DXF.
Signing off.