What are known extreme examples of protein translation/effect coupling/decoupling? For example, examples of proteins that are immediately used at the time the have been translated and vice versa, cases of proteins that are translated but then only used much later in time in/outside the cell?
Proteins that are membrane localized or secreted would take time to become active or even become poised for activity. Membrane receptors may remain inactive until triggered by the appropriate ligand. Nonetheless proteins won't remain (inactive or active) for a very long time as constant turnover happens in the cell.
One example, similar to what MarchHo mentioned in their answer (pre-proteins) is that of fibrinogen. Fibrinogen gets converted to fibrin when there is an injury and haemostasis reaction cascade is initiated. Fibrinogen has a biological half life (in its latency) of ~100 hours [ref].
Chymotrypsinogen is a proteolytic protein that is translated and stored in the pancreas in an inactive form. They are only activated when they are secreted from the pancreas.
Other zymogens such as apoptotic caspases have similar lengths of time between translation and use.
One example that may interest you Rory M are the so-called toxin - antitoxin pairs in bacteria. Interestingly enough, many bacterial species need a way to protect themselves from their own toxins. Thus, for a particular toxin, a corresponding antitoxin is synthesized (most often both are proteins, but the antitoxin may be RNA as well, three types of toxin - antitoxin systems are known; check out http://en.wikipedia.org/wiki/Toxin-antitoxin_system). The antitoxin remains bound to the toxin until the latter is secreted into its designated location. If the antitoxin decouples prematurely, cellular death may occur. I'm not too sure how long the two are "stored" within the cell; I imagine this could vary greatly between species.