Experts are valuable partly because they know which mistakes have already been made.
Spend ten years in a field and you develop shortcuts that a newcomer does not have. You can look at a problem and quickly tell the genuinely difficult part from the part that only looks difficult because somebody is new. That kind of judgement saves enormous amounts of time.
The cost of fluency is that some assumptions disappear from view.
A rule may have begun as a sensible response to a real constraint. The technology changes, the market changes, the customer changes, but the rule survives. A few years later nobody can quite explain why it exists. People only know that this is how serious people do things.
An outsider is often willing to ask the embarrassing question. Why does the customer need seven steps? Why is this sold through an intermediary? Why does approval take three weeks? Why is a limitation being treated like a law of nature when it may simply be a habit with good branding?
Most questions like these do not lead to breakthroughs. Sometimes the outsider is simply missing something obvious. Real constraints exist, and experts usually know many of them for good reasons. The advantage is not ignorance. It is the chance to notice which parts of a system are unavoidable and which parts have merely become familiar.
People arriving from adjacent fields can be especially useful because they bring working habits with them. Something ordinary in one industry can look radical in another. Often a new idea is not new at all. It is an old idea that has finally crossed the right border.
Status matters as well. Insiders have careers and reputations tied to the current system. Being wrong can threaten more than a proposal. It can threaten standing. An outsider may have less to protect, which makes it easier to stay with an unpopular question for another week or another year.
I do not find the lone-amateur-versus-the-experts story very convincing. The better pattern is usually a partnership: somebody willing to question the inherited assumptions, and somebody experienced enough to tell a real opening from a misunderstanding.
Good organisations make room for that combination. They ask new employees what looks strange before everything starts looking normal. They bring people from neighbouring disciplines into important problems. They let unconventional ideas be tested seriously, including against evidence that may kill them.
Expertise should make a field easier to navigate. It should not turn the field into private property.
The interesting work often starts where somebody knows enough to respect reality but still feels free to ask why the route has to look exactly like this.