No, this is not necessarily the case asthe results are subject to some variation based on probability. This variation approximates a Bell Curve, and with only one data point it is not possible to know where on the Bell Curve the successful result falls. It may be that the design is easy and that it will always meet timing, or it may be a very challenging design that will often fail as the design varies and the successful run represents a rare best-case result. The test for this is to run multiple Cost Tables (also known asvariability passes) and note the frequency of success. The variability pass change has the same effect as a minor design change and can be used to judge how repeatable the timing result will be going forward. The goal is to have the design performance goal fall as far to the left on the Bell Curve as possible so that the majority of passes are successful. To do this, the design should be tuned until a majority of variability passes succeed, not just a single pass.
See (Xilinx Answer 35534) for moreon Cost Tables and the Bell Curve.
Note: SmartXplorer can be used to find the best strategy before testing using variability passes.