I feel the round maps in PA is it's best and strongest suit, I don't know what any of you have against it. There's great room for innovative strategy and overall good RTS gameplay there all on it's own.
It's just comes with it's own set of design and mechanics challenges, mostly when it comes to how to synthesize the map data and provide an overview. Cola_colin came damn near to the best solution, his work should have been expanded on though and refined so as to provide more meaningful data. I had an idea but no means to implement it.
I do agree PA is nor a successor to FA nor a better game.
sadly it is inferior in all points left for :
- 64 bit
- linux support
- multi-threading support
- round maps
- AI (god that AI is so good)
I would have counted the server-client architecture as well if it wasn't for having coupled that with some of the most abysmally underpowered servers in proportion to the task at hand in gaming history.
In the end PA's fault's all come down to lack of budget. You can't blame them for not listening, they do more than any other developer i know, but it's obvious they could have gotten twice as much on the kickstarter, with twice as much actual benefit from it had it been set up more ambitiously.
the success bar was way to low and the rewards worth way too much actual money. (almost as if they were handing it all back robin-hood style. I'm not really persuaded the higher tiers generated any money.)
capitalizing further on SMMC paid content before launching the campaign and waiting for a better time might also have been more clever.
in the end the code that uber generated is surprising you almost all of the time. it's elegant.