this sort of thing has always been potentially complicated because the taxon of any given observation often changes after the observation is submitted, and observations aren’t always submitted to the platform as soon as they are recorded in the field. in other words, you’re dealing with 3 dates (observed, submitted, identified), plus the snapshot / retrieval date of the dataset itself.
it looks like you’re framing your view of the numbers only on observation date, based on the dataset as it exists right now. so if you’re thinking about things from that approach, i think the most useful way to look at the sort of data you’re wanting is using @glauber735’s lifelist timeline tool: https://glauberramos.github.io/inat/lifelist?place_id=186491&place=London%2C+Ontario
note: i think @glauber735’s tool works similarly to how @kildor’s tool (mentioned in the previous post) works, but i remember there being some lingering issues in @kildor’s tool in specific cases.
there are other ways to approach the question of finding “new” species, of course, but what’s most useful just depends on what exactly you’re trying to do at the end of the day.