Just thinking about this second exposure thing.
Most of the threads discussing "conventional" film reversal processing imply that the secondary exposure length is not that critical, as long as one does not try to use the sun or other UV-rich source and that all areas of the film, front and back, receive sufficient exposure. Apparently, extreme amounts of secondary exposure can push the unbleached areas into the solarisation region, which will produce a reversal of areas of extreme overexposure (black sun in landscapes for example). This is different from the Sabatier effect, caused by moderate light exposure while development is taking place, which can cause a partial reversal of the entire image.
It seems to me that the initial exposure is going to result in regions with varying levels of exposed silver halide, depending upon the density of that part of the scene. If the first development is to completion, virtually all of the exposed silver grains will be reduced to metallic silver. Only those grains will be (or should be) bleached by whatever chemical is being used. What remains are grains of unexposed silver halide.
My reasoning is that the overall density of the image is set by the initial exposure and development. One cannot over-bleach an image (ignoring other issues like emulsion softening, blistering, etc.) as the areas that can be bleached have already been determined by the first exposure and development. Likewise, one can not over-expose during the second exposure (nor over-develop during the second development), as it is desirable that all grains of newly-exposed and unbleached silver halide be converted to black.
In my thought model, each grain or crystal of silver halide has a binary state, black or not black, no in-between. Gray variations are caused by the density of the black particles in a region, rather than variations in the light absorption of the individual grains. I think this is correct, please feel free to jump in.
If the hydrogen peroxide bleaching leaves behind insoluble secondary actinic or light-sensitive compounds as part of the process, all bets are off and the newly formed light-sensitive compounds not related to the original exposure could cause all sorts of image defects, like the coppery stains we are seeing. It has already been stated here by those in-the-know that silver citrate, the product of the peroxide bleach with citric acid, it relatively insoluble and remains in the paper until fixing. It is also slightly light sensitive and may react with second exposure and development.
The silver citrate produced by the peroxide/citric acid bleaching is soluble at 28.4 milligrams/liter in water. The silver sulfate produced by the sulfuric acid in the dichromate bleach is soluble at 8300 mg/liter in water. It sounds like more of the silver citrate is remaining in the paper after/during bleaching, which may account for image issues during the second exposure and development.
Don