Before I try to cut down the incubation times for antibodies, I make sure that the antibody works in my hands in my regular protocol. I may have misunderstood your question, but have you gotten the primary antibody to work in your regular IHC protocol on the acetone-fixed sections? One thought is that the acetone might be changing the epitope just enough to prevent the primary antibody from binding. Although it is probably not compatible with LCM, you can test this by fixing a few of your frozen sections for a short time in 4% paraformaldehyde, and then running the IHCl reaction (the same as you are using with the acetone fixed material).
The uneven labeling you see suggests to me that perhaps the sections are not getting well fixed with only two minutes of acetone, especially if they are thicker than 10 um. You might need to increase the time to 5 or 10 minutes, since you then want to put them in a lot of Triton-X 100.
I don't know a lot about LCM. Is there some reason that you need to use DAB labeling, rather than a fluorescently-tagged secondary (the Arcturus system appears compatable with Cy-3)? I think that it would be easier, not to mention faster, than the HRP-DAB system. It can be difficult to get a good level of DAB reaction product if the reaction is run for only 5 minutes and there are low levels of HRP.
I forgot to add that if you do need to use DAB, one way to make the reaction product darker is to run all of your IHC steps using Tris buffer or TBS rather than phosphate buffer or PBS. In fact, you can just make the DAB in Tris buffer if necessary, although in my opinion that doesn’t result in as dark a reaction product as using Tris during the entire procedure.