Is someone able to explain to me the process for achieving RG status for an observation? Is the opinion of two or three inexperienced identifiers treated exactly like an ID offered by experienced/expert identifiers? Do curators have a special status in identifying observations or are they the same as any other user?
Do curators have any ability to over-ride incorrect identifications by inexperienced users?
Everybody’s IDs are treated the same, experts/curators have no special abilities here. Anyone can add an agreeing ID to the initial ID by the observer to get something to RG, and also anyone can add a disagreeing ID to bump it back if it seems wrong.
How in-depth do you want your answers? And why do you ask?
Process for achieving RG.
The observation must fulfill a list of requirements, including that it be wild, have accurate geopositional data, be of one specific organism, have a photo or sound, etc.
There are a number of citizen science platforms such as eBird that allow the opinion of a single person to be enough for the data to be added to databases. Others may be far more strict. iNat takes a middle path: more than 2/3 of the identifiers must agree on a taxon, and that taxon must be specific (or, below tribe and marked as impossible to narrow down further).
This means that a single disagreeing ID on something previously RG requires another, fourth person to step in.
Status of identifiers.
Yes, iNaturalist allows all users to make identifications in good faith. It explicitly asks users with less experience to be cautious in adding IDs, to ask questions of those who know more, and to be ready to withdraw or correct themselves when presented with evidence.
iNaturalist chooses to recognize expertise from within the system, rather than through outside factors – a degree may have a high correlation with ID accuracy, but it’s not democratic to forbid citzen scientists (and this is a platform for citizen scientists!) from making their own best attempts at identification.
Curators do not have any additional voting power. It is a role of added responsibility, not one of added power.
Any user may over-ride incorrect IDs by adding their own ID. But you get one vote per observation. You can always confer with other users who are more experienced, but it is always preferable to teach the less experienced rather than berate them.
If the observation is one of your own, you have the power to unilaterally reject the IDs others have put on it and it will be classified as the taxon of your choice. However, if this means that the ID consensus and your ID conflict, it will not be RG. A small number of people use this option.
iNaturalist produces a huge amount of vital biodiversity data, but it is a platform meant to serve as outreach, education, and socialization of amateur scientists and students. The quality of its data has been measured as comparable to those of many museums, and it does this while helping answer the curiosity of its users about the world around them.