I am a chemist, so I can make some comments on this.
Household bleach is a solution of sodium hypochlorite (and other things, such as sodium hydroxide) in water. The active ingredient is the sodium hypochlorite, which is a salt that most certainly does NOT evaporate. Evaporation of the water would leave the solid salt behind, so traces of oxidising agent can be there for a considerable period of time (especially in small cracks and pits in the tile surface) even though the liquid is gone.
There are also solid bleaches based on calcium hypochlorite available.
Luminol does not react directly with blood. The fluorescence is produced by mixing it with an oxidizing agent (which is why bleaches interfere). Investigators use hydrogen peroxide as the oxidising agent.
Certain transition metal ions, such as iron or copper, act as catalysts for speeding up the oxidation of luminol by the hydrogen peroxide, which is why traces of blood (which contains iron) makes it glow brighter compared to surroundings.
NOTE: Blood is not the only source of iron in a household, anything that can rust will generate a positive response too. Copper and iron are
very common in household objects. These metals are also very reactive to oxygen when exposed to air, consequently any metal surface is going to be covered by metal salts and these salts can be dissolved if those surfaces are exposed to water. For example, your mop is probably secured by copper or iron pins, so using it will spread that around. If your feet were wet and you stood on a metal surface, then walked around, you would potentially leave footprints containing traces of these salts behind. Other redox metals such as manganese and chromium probably catalyse the oxidation reaction as well, but those are less common in a household environment.
The point is that these luminol positive spots could have come from any number of sources which are commonly found in households, they do not have to come from blood. To show that your luminol positive areas contain blood you have to do secondary tests (probably for haemoglobin or some other erythrocyte specific protein) to show that blood actually is there. My understanding is that these secondary tests were either not done, or were negative. They get around that by saying that sometimes a blood stain will not test positive for blood in the secondary test. HOWEVER, if the secondary test is negative you cannot conclude that blood is there, only that it might be there.
You can read more about luminol testing here:
Luminol - Wikipedia, the free encyclopedia
Note: according to the Wikipedia article, one of the drawbacks of luminol is that fecal matter will produce the same response as blood and it is a potential false positive. Fecal matter is another substance that could have been deposited as a "footprint" at some point in the building's lifetime.
Getting on the DNA results for these "bloody footprints". This DNA would had to have come from cells. Blood contains (for women) about 4.5 million erythrocytes per mL. If AK left DNA in these footprints it would have to be from shed skin cells from her feet. Any such shed cells would be far fewer than 4.5 million, likely not much more than a few hundred at most. Assuming there was a smaller quantity of "blood", let us say 0.1 mL (which is about a drop), that would produce about 450000 MK cells mixed with the few hundred from AK. That means that there would be at least a 1000 times more MK DNA there than AK's, and you would think that the MK DNA would drown out any sign of AK DNA. The chances of detecting AK's DNA, but not MK's, in these "footprints" therefore would be at least 1000 to 1, probably more.
Every "footprint" that shows AK DNA but not MK DNA would have the same odds, and the overall probability would be the product of those numbers. So, if there were, say, three footprints like that, the odds against it being AK tracking MK's blood would be at least 1000000000 to one (a billion to one).
According to that wiki page you cited there were three "footprints" with only AK DNA, and one with both AK and MK DNA. But, the odds that they would only find AK DNA in so many footprints is astronomically unlikely, so large in fact that it argues convincingly against the prosecution's argument.
The DNA they found is probably from skin cells sloughing off normally from AK and MK that just happened to be in the area of the "footprints", and since they both lived there it is not surprising to find it.
IMO these luminol positive areas they found were either not blood, or had no relationship with the crime. Certainly the evidence would suggest that.