That is the main thing about citizen science and citizen science based initiatives like iNaturalist: it is gathered and some times confirmed by non-professionals.
I know there are a lot of professional taxonomists using iNaturalist (same for zoologists, botanists enthomologists, etc etc). But most of the times, your observations will be confirmed by a non-professional.
iNaturalist “shields” itself from bad indentifications by requiring a consensus: if the majority of poeple don’t agree with your ID, it won’t become research grade.
The problem with that is that the observer (lets say you) can miss-identify the observed organism in the first place. From this situations, 2 things can occur:
#1 - A watchfull identifier notices the miss identification and corrects it. If another identifier agrees with the first identifier, it reaches research grade with a correct identification.
#2 - The identifier agrees with your miss-identification (either because he/she is simply agreeing with everything - yes, it happes… - or because he/she consciously agrees with your bad miss-identification), and it reaches research grade with a wrong ID. In this case, we just hope someone finds it and does what is described in point #1
To sum it all, you can’t assume everything here will be indisputably correct. You can assume that it will be the closest to correct as it can, given that people follow a good ID policy
(sorry for the long answer and for some possible typos, english is not my native language)