What I've learned from here over the years is that the best method/approach to forecasting is to account for everything...especially flags. When flags start to arise, that's when you need to pump the brakes and start asking yourself questions in your head...questions such as "how does this alter the potential outcome" or "what does this offset". What I then do is try and paint a visualization in my head of how everything will look as the event is unfolding (especially regarding radar).
This does burn me often b/c I will be conservative more times than not, however, I have learned that there are some things I may place too much emphasis in or that my understanding of what I was interpreting was just flat out wrong. One example of this is RH in the DGZ. For some reason I was using 90+% as a threshold, however, upon reading up, I was way too high...threshold seems to be more in the 80-85% range...so when I would see less than 90% I would think it was too dry for good snowgrowth and I would underforecast.
I also suck at not necessarily analyzing charts like 500 vorticity, but I really suck at understanding how the pieces work together and how subtle changes in placement, structure, track influence everything at the surface. This BURNED me last year...remember the weekend where we had a storm Friday and sunday or Monday...I completely played up the Friday one and and downplayed the second one...we got little Friday and slammed Sunday.
That certainly presents a major challenge for you...and everyone else really. It has to be tough when forecasting along the boundaries of CWA's in questionable situations...you think 1-2'' but just across the border they're going like 4-6''...huge difference and that does the public no good. People really thought you were ignoring southern NH...lol