If I could have a week with the code, I could turn this into a 9/10 game.
You play the game by placing flags and bounties, and your heroes respond based on the risk and bounty of the flag. Sounds like a good idea? Well, it is but poorly implemented. Your heroes are cowards when they should be brave - if your castle is being attacked (you lose if it is destroyed), they run away unless you have a lot of gold. And your low level 1 warriors fearlessly charge lightning elementals and are slaughtered. Your castle is being attacked, but not enough gold to put a bounty on it? Your heroes just wander aimlessly 20 feet beside, "looking for adventure."
In combat, the AI is suicidal. Low on health and without health potions, you'll watch in dismay as your high level hero chooses to die in combat rather than run away and heal (which takes all of five seconds) or buy more potions. You can resurrect your heroes, but it's bizarrely expensive, so if your hero is mid level or above, you won't be able to afford to bring them back. You can put your units into parties - that is, have a couple of warriors, a ranger and a cleric all work together as a team, but again, this is so expensive that you can't use it until later in the game when it's much less useful.
Playing the campaign is mostly an issue of building archers, sending them in all directions to find where the enemy spawns and trader points are, restarting and doing this a few times, and then hoping that you get lucky. In every map, you are attacked from all sides, and with the frustrating controls it just adds to the pain.
What bothers me the most about the game is that they were so close - a few simple AI routines, tweak some numbers to make resurrections cheaper and AI more responsive toward flags/better at prioritizing, and the game would have been fantastic. Given a week to tweak the game, and I (or anyone who's played it, really) could make it fantastic.