No. I said that for some taxa in some regions this is the case.
There are also other factors at play in your graph, such as the growth and expansion of iNat (more people making observations in more places) that mean the growth curve does not show the same pattern as it would if the number of observations were consistent over time. I imagine if you were to control for the number of observations this graph would look very different – note that the change in the number of species over time in your chart appears roughly linear while the change in number of new observations per year on iNat is more exponential. I expect that the ratio of of new species to observations is likely broadly decreasing – it would be highly surprising if this were not the case, for all the reasons I gave above, regardless of how difficult or well studied the taxon in question is.