Transparency. A question of experience, but also a problem of organization. A project often fails - completely or partially - because the PM (or a Team Leader) does not act in a transparent fashion; in other words, certain project limits, analysis limits or an execution-related mistake are not disclosed in the hope that these problems will fix themselves or be resolved by some form of divine intervention. I like to say that this is a problem of organization.
The Project Manager truly is one of the main factors responsible for the success of a project, but not the only one. A Project Manager cannot magically transform incomplete analyses by specialists into valid and exhaustive reports, neither is a Project Manager capable of verifying every technical product at every stage of the project. The fact is that the PM is usually the architect of the project, is emotionally involved and considers the project as something of their own creation.
The PM is often an employee of the company that launched the project, whose opportunity is inversely proportional to the project cost. Or they believe that a good PM should be able to make something that does not work, work. For one or more of these reasons, the PM starts to instinctively adopt what can be somewhat dangerous attitudes during the analysis stage, tending to maximize the pros and minimize the cons in each case. The Project Manager should act transparently for the organization with an interest in the project and especially with the sponsor, all the more so if the latter is known for being a “killer”. The PM should immediately report any significant problems, each flaw, even if they are directly responsible - obviously without becoming overly anxious - and share any relevant question or unforeseen eventuality with the stakeholders; not doing so could lead to very serious, possibly irreparable, consequences. Hence, generally-speaking, the sponsor should choose a PM who is not emotionally involved in the project. A Project Manager who always acts honestly and entirely transparently has great potential to be highly appreciated, even when a project fails.