The problem I have with DotA isn't the elitist environment but the awful user friendliness. As a general rule, those people are elitist for a reason and I don't really have a problem with that. What I don't like is that I feel like I have to go to school to figure out how to play the game. Coming from a fan of the notoriously complicated X Universe series, that's saying something.
Anyway, I do agree with many of your points. The War Rank and Gold all come fairly linearly and there is little you can do to get ahead in the current balance. The War Rank slowly increases over time at the same speed as the enemy and gold is obtained through mines and killing Demigods. I don't think either of these factors is really a bad thing, but they sound more like part of the foundation than the full gameplay. Unfortunately, that is currently the full gameplay.
So what can we do about it? You specifically mentioned changing how towers worked. I agree that something should be done here. Many times I have pushed for a tower through a very difficult battle and felt cheated when it finally fell. I mean, sure, it isn't shooing at me anymore, but I didn't feel like I had accomplished something and really aided the team. I get to go forward a little, but I'm not going to get to another flag or make a meaningful attack on their base, so how does that extra plot of land help me? You could give people a large gold bonus for killing a tower, but that just sounds rather bland. More on this in a moment.
I think the War Rank is currently downright wrong. The fundamental system is flawed. Both teams progress at a set rate through a series of ten ranks that are always equal. What's the point? If you're going to be lame about it, why don't you just make each creep wave spawn with one more guy or something? I propose turning your War Rank into a full fledged secondary currency next to gold. It would be used for all Citadel purchases (possibly in addition to gold) and its current job of upgrading creep waves. Gold would be for each individual Demigod, and War Rank would be for the team as a whole. I'm not sure if it should be split among the Demigods or pooled together, but I think the concept has potential.
This would have several interesting effects. Most notably would be that the teams wouldn't be side by side in troop improvements. Things like Catapultasaurus rushing, Giant rushing, and Minotaur spamming would be introduced. It would enable a variety of different team strategies beyond just tactical point-and-click abilities. You could tech (focus on gold production), turtle (focus on tower upgrades), or rush (focus on creeps). You could go for quickly attempting to get large units onto the field, or you could focus on improving the smaller units you have before expanding upward. Obviously, there would always be the balanced path in addition to the extremes. Those precious catapultasaurii would be a risky venture instead of a surefire thing, as they could be balanced to be glass cannons. They won't really help you at all unless you're actually pushing against the towers, so if you need to be securing flags and territory you would be better off to buy improvements for your Minotaurs. Delay too long and they'll start stomping you with giants. You would have to make strategic decisions in order to triumph.
But this still doesn't solve the problem of lame and boring War Rank. It still only goes up with time, and even then at the same rate as the other guy. You could always make an upgrade for it or add flags, but that's not really special. Now I would like to go back to the towers. These are not symbols of other Demigods that you attack and destroy, but a defender of the entire other team. Doesn't it follow that by attacking something that represents the enemy team as a whole you should be rewarded by gaining something that aids your team as a whole? I propose that War Rank would be primarily decided by how much territory you captured and how many enemy towers you claimed. Destroying a tower before catapultasaurii appeared would actually aid your team's struggle. It would put a value on the fairly unimportant exterior towers and it would give you targets other than the Citadel to strike at if you make an assault on the enemy base. Currently it's all or nothing, but having smaller targets would make base assaults useful at all stages of the game.
Imagine if, for example, every tower you destroyed would allow you to purchase a citadel upgrade. That early scuffle wouldn't just be for the experience, but you would really be gunning for the enemy tower. If you can take out that forward tower, you can upgrade to including Archers in your forces. Isn't that a worthwhile goal? Obviously just a rough example, but I hope you follow what I'm saying.
The one obvious flaw with everything in this thread is what DeadMG pointed out about the slippery slope. This is present in almost all games, and there is very little you can do about it. Everything in here makes comebacks harder. This isn't necessarily a bad thing, the winners should win and the losers should lose. However, many people have a problem with games that are decided in the first few minutes and then drag on for half an hour. The whole concept is slightly paradoxical. On one hand, the winning team should be rewarded for playing better and have their effective actions yield relative results. On the other hand, you don't want one or two effective moments to give one team a decisive advantage over the other. This really just comes down to balance. The slope will be there in every game, but how deadly it is depends on how well things are balanced. I think that once an effective system is implemented, it can be modified so that the slippery slope is slightly less steep. However, at the moment we lack an effective system, so developing one should be the first priority.
I am curious to hear what others think about this post. Also, Mario, what exactly do you propose we do about everything? I honestly just thought this up on the spot, so I fully expect it to be full of flaws
. Perhaps, at the very least, it can be fuel for practical discussion.