The accuracy of estimating project costs relies on just five concepts. Strangely, detailed plans of the proposed building are not necessary - what is necessary is as much inclusiveness as possible. The accuracy that matters is the bottom line. Individual line items only matter to the contractor performing the work.
Concept One: Size
Size is fundamental to estimating cost. Most, but not all costs are related to size. Normally lack of space is what drives a project. Quantify how much space would solve the problem and you have the basis for estimating project costs.
Concept Two: Breakdown
The more line items you have in your estimate the more accurate your estimate will be. This is really just the mathematics of the problem expressing itself. For instance if you have just three line items - construction, design, and soft costs - you cannot be far off on any one of them and still hope for an accurate estimate, say less than 5% too high or too low. On the other hand if you have 200 line items, each one is significantly smaller than the first example; and you can be too high or too low by quite a bit on any line item without major impact - and your errors elsewhere will likely offset one another.
Concept Three: Research
If you don't have any idea how much something will cost, find out more about it. Blind guessing will eventually overwhelm the Breakdown Concept.
Concept Four: Bracketing
This concept comes from field artillery. The first shot isn't expected to hit the target - it is a range finder. The second shot corrects for the first shot, and in turn adds critical information for hitting the target. The third shot is the hit. When you are estimating project costs, the first "shot" is what you think the minimum cost will be. The second "shot" is what you think the maximum cost will be. The line item cost that you go with is the average of these two extremes.
Concept Five: Contingency
You always know more about your project at the end than at the beginning. You compensate for this lack of knowledge by using a contingency that is sized for the level of information that you have. If you don't have drawings yet, use 15%-20%. At each succeeding phase reduce the contingency until you have bids in hand. Even then 2%-3% for a large project is wise to budget for problems during construction. Small projects are very hard to estimate accurately, but a larger contingency will keep you from being surprised by higher than expected costs.
While there really isn't a better or more accurate way of estimating project costs than outlined above, an interesting thing about estimating project costs is that early in the planning for the project you can steer the costs dramatically in the direction that you want. If you use those line item estimates to guide your decision-making, it is amazing how close you can come to the target that you set for the project.