It seems (from the General's perspective) that this blind auction/draft system was suboptimal. I have a somewhat simple suggestion that could improve this system for future wars (if it isn't completely overhauled).
Current System: (Highest Bid General with Open Roster Spot) gets (Player X) for (Highest Bid on Player X made by General with an Open Roster Spot).
Proposed change (Highest Bid General with Open Roster Spot) gets (Player X) for (2nd Highest Bid on Player X made by a General with an Open Roster Spot)+1.
For example:
InsignificantWeaboo really wants DoubleCapitals on his team. In a blind auction that means he needs to bid pretty high to have confidence he will get DoubleCapitals. To make sure, InsignificantWeaboo bids the max 7500. However, other generals do not rank DoubleCapitals as highly, and the next highest bid by a General with an open roster spot is 1002 by ShockCannon. Because InsignificantWeaboo bid the most on DoubleCapitals, he receives DoubleCapitals onto his team, but because the next highest bid was 1002, it only costs InsignificantWeaboo 1003 points (not 7500).
This allows the general that bids the most on a player to get that player, but the cost is determined at a "closer to market" rate rather than a blind guess by a single general.
Generals who prioritize a player or two would get them at whatever other Generals essentially valued those players at. If two generals really want a player, that player will cost much more, but in many cases this helps alleviate the price hike that comes from the blind nature of the auction.
I understand there is a certain amount of strategy to be employed in the current blind auction system, and some generals probably quite enjoy that. This proposal changes the nature of such strategies, but I am certain it still leaves quite a lot of potential for strategic bidding.