When I created a mod inspired by DyP, one that started in the industrial age, I noticed that the high number of improvements I made available, prevented the AI from competing adequately. Production-boosting improvements were always low on the AI's priority list, and they were necessary for the AI to be able to compete adequately in the production of high-cost units and improvements. I know that RAR has tried to address this somewhat by raising the shields provided by terrain and bonuses, and I've seen the comment that the AI competes better when a 3 billion-year-old world is chosen. I wonder if simply reducing the production bonuses of certain improvements may also improve game balancing.
I've seen the comment several times by developers of the mod that it is designed so that a player who tries to build everything will not succeed; players instead must carefully choose what to research and build based on goals and priorities. I appreciate this approach and it meshes with my game philosophy: if there is only one best way to do things, the game becomes meaningless. This is great in multiplayer games, but against the AI it is crippling for your opponent. The AI really does try to build everything, and doesn't choose very well what things to build first that will lower production costs so that everything can be built faster.
My recommendation would be to lower the cost of everything that has an inflated cost associated with assumed availability of production-boosting improvements. As I examined the unit and improvement costs in DyP (and I assume that RnR has a similar system), I noted that costs increased as more production-boosting improvements became available. Essentially this forces a player to research and build the production boosters (Forge, Water Mill, Factory, etc.), essentially undermining your goal of giving the player multiple pathways to victory. In a sense, this forces at least one pathway, the production pathway. Unfortunately, it is not one that the AI understands, and leads to a less balanced game.
I've also noticed in these threads that the AI's neglect of air and naval forces is a disappointment to many players. I've seen the development team's explanations, and I accept the reasons for this phenomenon. I have seen the same thing in regular C3C games I've played, where the AI will scramble to catch up in ground power before switching its focus to air power, and naval power is always neglected, except by some seafaring nations.
Perhaps one answer to this would be to address the issue in the civ preferences in the editor. Perhaps air and naval units could be flagged as a build preference for ALL civs, and perhaps even the ground forces preferences could be unflagged even when it would not fit with the realistic flavor of a civ. I'm working from the assumption that there is a built-in preference for these units, and that flagging these in the civ preferences page will just double it up. Faulty assumption perhaps....
The Banker Phenomenon has been addressed as one of the reasons for ancient-era units surviving into the Modern Age. One suggestion I have here is to make a Leonardo's Workshop-like small wonder that all civs could build in the industrial era. I also favor more Admiral of the Fleet-type units, and wonder if it would be possible to have a small wonder associated with *every* unit in the game. The idea would be to have it be an add-on to some other building, such as a barracks or air base (for ground or air units, for example), and could even be more specific, such as a Castle add-on for mounted units in the middle ages. The idea would be to have the add-on function for a limited amount of time, becoming obsolete before it had made *too* many units, at a historically-appropriate time. This would also free up the AI from needing to build so many ground units by hand (to meet its irrational need to match you in ground power), so that "exotic" units (air and sea) could be built.
I should add that although I played DyP a lot, I'm just starting my first RnR game. My brother and I play a lot of PBEM games, and we're waiting until the next big RnR update before using it for PBEM. Our games tend to last 3-5 months, and so we want to make sure the mod is well-refined before putting so much of a time investment into it. But we're both very excited about it, and I want to give a big thanks to everyone who is working on the bugs, civilopedia editing, and so forth. It is greatly appreciated!
.
