I'll say what i've said several times before (and been ignored!):
1) You can NOT have ties broken by who attacked first/second as this is a worldwide game with all the associated Time Zones so some people will be asleep when GAC Attack rounds start, or at work etc.
2) Breaking a tie based on who has the higher/lower gp is an awful way to deal with a tie. There are multiple "performance" based metrics that could be used, so to just award it to who has the most (or least) gp just feels utterly wrong. However if a sensible person were going to use this particular value then they'd award it to the person with the lesser gp as a way of potentially recognizing "stepping up" against a larger player
3) The way a tie SHOULD be broken is by a "performance metric". Personally I'd do it by awarding the win to whichever player had the most 64-banner first-time wins, if still tied then the most 63 and so on. Then if still completely tied go to whoever had the most first-time undersized squad / fleet victories. Then if still tied go to whoever had the most Defensive holds, then if still tied who had the most Offensive first-time wins, and if still completely tied (unlikely!!) Only at that point would you move to using gp - I'm sure there are other performance metrics that could be totted up and used, or people may prefer to use the ones i've mentioned but in a different order, but pretty much anything along those lines would be better than just "big gp wins"
4) There's no reason why the points or new GAC store credits couldn't simply be split equally while there is no performance metric being used to determine the winner!