I thought that it may be useful to gather a collection of images together that are recorded as having major flare events associated with them for comparison. Dual polarity separation lines, intense (and closely connected) patterns and a nucleus of one polarity closely bordered (almost surrounded) by regions with the opposite polarity are some of the more common elements. We only see a snapshot of a moment in time and don't get too much information of how these things evolve over time but there may be some similarities...
I've been using #major when I've come across these if you can get search to work properly. Any comments?
I think that is a great idea. As we've talked about before, many times people can pick out patterns that automated methods have a hard time seeing. Finding a pattern that always results in a major flare is definitely a holy grail of solar flare prediction research. Of course as you allude to, the patterns may be in the evolution of the sunspot group.
Some times the question is 'the sunspot group had all of these flaring patterns present, so why didn't it flare?!'.