Analysis

Free land gives 10 gold per hour. Mines cost 500 gold, then give 30 gold per hour. So it takes 25 hours (about a day) for mines to pay for themselves, after that they give 20 gold per hour profit.
The cost of other buildings is not mainly the 500 gold required for purchase, but the lost opportunity of mine income. Hence, any other building costs 30 gold per hour.
There is no reason to have free land, except that you don't have enough money to develop it yet. If you have the money to do so, you should be buying mines as soon as possible, to maximize your income. Alternatively, you should be developing your land to gain the benefits provided by other buildings as soon as possible.
The best way to think about development strategy is in percentages. All free land should be turned into buildings, the question is, What percentage of land should be used for each building type?
The simplest (and also arguably the best) strategy is 100% mines. Other strategies can be compared with this simple benchmark.
Two important constraints help us narrow down the number of optimal strategy alternatives. The first is obvious, we cannot have buildings covering more than 100% of our land. The second is that there is no point having more than 30% forts, for example, because the maximum defensive modifier is 45%, and it only takes 30% forts to reach this level.
For purposes of comparison, we can see that maintaining a 45% defense modifier costs 30% of your maximum possible income, because 30% of your land is used for defence rather than income.

Comparisons

Mine income is particularly important because it is this regular income that is needed to maintain your army. A standing army cannot grow beyond the income you have to support it. Details of efficient army structure will be covered in another section. For simplicity, let's assume all armies are composed only of soldiers (which cost 1 gold to maintain).
Giving up 30% of your income means giving up 30% of your standing army, which is fine, because 70% (the remaining army) gets multiplied by 1.45% (the modifier for the forts) giving 101.5%. By buying forts instead of mines, you are 1.5% stronger in defence. However, you are only 70% of your previous attack strength!
This is not quite as bad as it seems, because the defence modifier works like a "surprise attack" -- players who don't take it into account when attacking you will lose, and you will gain experience.
The numbers work exactly the same way for training grounds. You can reduce your defence to 70%, to boost attack by 1.5%. It is difficult to see why anyone would do this however. All attacks are surprises in WarBook, usually on weaker players.

Magic Buildings

The other two kinds of buildings cost only 25% of maximum income for maximum effect. Amplifiers will add 50% to your spell damage, and barriers will give you 75% protection from spell damage, if you have the 25% building coverage.
Generally speaking, the aim in WarBook is to increase one's own kingdom, not to damage others' kingdoms. Assets spent on damaging other players may damage them more than it costs you, but players that invest in building their own kingdoms rather than damaging others will move ahead more efficiently. As pure damage maximizers, unless they are increasing land grabs (magicians only), or reducing a target for conquest (an expensive option), they are of more use in wasting your resources and in gaining you enemies for a fun, but very short lived, WarBook experience.
On the other hand, magic attacks, if used, can be devastating. With full barriers, a magic attack will be reduced to 25% of its normal damage. With no barriers, magic attacks are four times more devastating -- the full 100%. The question is, Is it worth permanent 25% loss of income, for a 75% reduction in magic attack damage? It is worth remembering that magic attacks are a wonderful retaliatory tool, but likewise armies can retaliate against magicians and the acreage range of military attacks is greater than spells.

Conclusions

Finally, by surrendering 25% income, you can reduce training costs by 50%, using barracks. This can be a useful investment if your hero type strategy makes heavy use of its elite units. However, it needs to be noted that the 100% mine strategy will give a 1/3 larger standing army or 1/3 higher income over a 75%-25% split.
So far we have only considered strategies with one or two types of building, and with maximum use of the non mine buildings. Obviously, other possibilities exist, but the logic for comparing them remains the same and, in summary, they all involve smaller, tougher armies against larger, less modified ones. Under the current program, the modifiers are not sufficiently high to justify their costs. The mine only strategy seems to be the most efficient, except in exceptional cases.

Another Idea

This is a strategy that, although a bit costly, is quite efficient. First off, build up 70% mines and 30% Forts. This allows you to maximize defense and income. When you are planning to attack, do so all at once. Save up as much money as possible. Then when planning an attack run, destroy all of your forts. With the building make the max amount of barracks (25%) to buy military units. Then demolish the the barracks and make training grounds (30%) to max out your attack power. Complete all of your attacks. Demolish the training grounds and build forts back up to 30%.