Re: Civ5's gold mechanics covering the same role as Civ4's maintenance:
That's not entirely correct. Remember, Civ4 maintenance worked off of gold, the same gold you needed to balance with science.
I'm talking about the mechanics' role in limiting expansion. Civ V's system, without distorting features such as Tradition policies and the National College (and other 'you must build one in each city' national wonders, which lead to design tension with the building maintenance system), leads to a similar system of slowed expansion.
In Civ5, you only really need to balance your budget, having an actual gold income is just a luxury. Maintenance in Civ4 was also heavily tied into civics: if you had a smaller empire, you could afford taking civics that had high upkeep because your empire wasn't generating that much upkeep in the first place.
A gold income in Civ IV was a luxury; aside from maintaining the cost of civics, once maintenance was covered gold didn't do anything. Getting 'gold huts' as rewards in Civ IV was pretty much useless, as was Wonder 'fail gold', without the civics that allowed you to actually spend gold. Civ V goes too far in the other direction - gold is everything and a positive income is much more than just a luxury. It buys research agreements, city-state favours if you're short of achievable quests, and quick buildings and settlers.
The one big outstanding flaw with the system is the AI's willingness to trade anything you have going for gold, and in particular GPT. Higher difficulties surely shouldn't give the AI bonuses that reward the player more than lower difficulties, but the AI gets gold bonuses and doesn't use the gold itself, making them easier to 'farm'.
Building maintenance is just a terrible idea, because it punishes you for investing into your cities: to this day, I still find it ludicrous that settling a new city in Civ5 will increase your gold income before it decreases it, since building maintenance is the only way cities can lose you gold in Civ5.
The idea of building maintenance is to promote specialisation in your cities, rather than building everything you can because, once the city's founded, everything is free. It also helps prevent tall empires from running away financially because they accrue more and more gold as they grow while never incurring maintenance costs. I'd agree that it doesn't work as it should in Civ V - too many buildings, especially early in the game, are required practically everywhere anyway, and if not required for active city development are required to meet national wonder requirements. But that's a flaw in implementation rather than in principle.
Social policies are very different though: they're irreversible, necessary, and potentially game-ruining.
Yes, they're different in all those regards, but that's an entirely different complaint (save that the opportunity cost is higher in Civ V than Civ IV precisely because they are irreversible). As far as the strategic element is concerned, the two systems require similar opportunity costs - taking one bonus forfeits the alternatives.
"Should you have the opportunity to fix a mistake you made earlier in the game?" is an interesting question - given the timeframe of Civ games and so the effort invested into them, the answer is probably yes. And it can certainly be argued that civics can help you adapt to new information; while not especially common in a Civ game framework, since the challenges tend to be static and known in advance, it's a feature strategic play should allow. But Civ V is hardly the first strategy game ever made that doesn't allow you to take back moves that later prove to be wrong.
By contrast, if you choose Liberty opener in Civ5 before realizing you need to go 4-city Tradition, not only is that Liberty opener a wasted policy point that you can never get back, but it also increases the culture cost of future policies by virtue of being unlocked.
This is a very significant problem with Civ V; the issue again is with implementation rather than the principle of policy trees. The designers made the choice to have a 'tall vs. wide' tree decision early in the game.
This is fundamentally flawed, because at the game stage where you're forced to choose you're working from limited information. You don't know how far you have the opportunity to expand, or much about the other civs and their expansion plans. You can know that you can build one city, and may be able to build three or four. You can't rely on having the ability to settle early and often enough to gain more benefit from Liberty than you will on Tradition, purely by virtue of not knowing much of the map or how many cities your rivals may have that you'll be in a position to capture early in the game.
Social Policies in Civ5 are double-downs, which means they are either vital in winning you a game or instrumental in losing you the game, with the player often not having a say in the matter. The reason people prefer Tradition over Liberty in Civ5 is not necessarily because 4-city Tradition is so much more competitive than Liberty, but because you are less likely to end up in a situation where going Tradition turns out to be a mistake.
Exactly so. There's a further implementation problem as well in the loathsome 'finisher' bonuses, particularly the faith-buy unlocks - the game wants to force you to complete each tree rather than transitioning into others that may reward you better in specific scenarios.
I've said this before in plenty of tall vs. wide discussions, but I believe the choice between tall vs. wide should not be one that the game design addresses. In a game where expanding does not make you flatout weaker for the rest of the game, people will want to play tall instead of wide for three reasons: they want a challenge, they want a unique spin on things (this includes immersion), or they dislike the tedious management that comes with having a large empire.
I only partly agree with this; Civ's fundamentally a single-player game so shouldn't, indeed, be in the business of making sure every option is equally balanced with one another. All that matters is that players can beat the AI doing both.
But equally, players should be rewarded for playing a diversity of strategies rather than having to be at a noticeably higher skill level or wanting an extra challenge to play tall - OCS can be a challenge; not feeling duty-bound to build a settler whenever you have a positive gold income and a city in danger of becoming unhealthy shouldn't be.
One thing I will say is that playing tall is ultimately just rather boring - with fewer cities to maintain you can build duplicates of everything in all of them, and with only a few building slots there are a lot of times when you are - as Civ V is commonly caricatured - just pressing "End Turn" repeatedly or shuffling units around the map more for something to do than because they need moving. You aren't going to war because you don't want extra territory. One thing it isn't is a recipe to escape tedium; all of my best Civ V games have been playing wide (and not just because now that, rather than playing tall, is the challenge).
For the people in the first group, letting wide play do its thing with giant hammer outputs and gold yields reinforces the challenge. For the people in the second group, giving tall play weird and wacky advantages is better than giving them advantages that help tall empires compete with wide empires on the yields front, since it makes the tall experience a lot more unique than just "SimCity and don't expand". For the people in the third group, addressing their primary concern by making wide play less tedious is ideal, tall play should not be the sole option for people who dislike tedious micromanagement: better worker automation, better scout automation, better citizen automation, possible production automation, these are all things that this group would like without you having to change your game's design (puppet cities are an amazing idea, for example, too bad their terrible AI ruins them).
All of this is right on the mark, but something no Civ game has yet attempted - I like the idea of giving tall play 'weird and wacky advantages' though I'm not sure how it would be implemented.