no, you're not alone! we've actually had this conversation as a team and with yc batchmates as we've gone through yc w21.
to really simplify what we do, we're a reporting tool for companies that take out large loans from banks and need to send updates to their lender every month.
i think there are a few questions contained in your original question, though. among them: has the low-hanging fruit of tech/software been picked, such that only esoteric or niche ideas (there was a slightly heated conversation about a calendar app the other day) get funded? has abundant VC funding created a culture of solutions looking for problems? etc.
my hot takes, in order of conviction:
a) "[tech] can only be understood backwards; but it must be [created] forwards" (apologies for the misappropriation) -- another way of stating PG's point that tech startups look like niche/low-value toys at first before expanding into larger areas. this is a bit tired, but like all good heuristics, even if you know it you still fall into the mental trap. so when i look around and start to see other yc w21 companies as "toys" or inscrutable or bizarre, i try to interrogate that belief. easier said than done!
b) this could be a symptom of spiraling complexity in software (and the saas ecosystem). all the new yc startups are building on top of a few generations of other yc startups (e.g., now that everyone has a CRM + ERP, what could you build on top of that?). i was previously at a yc-backed enterprise saas company whose clients included many other yc-backed enterprise saas companies. my brother, a debt investor at a bank, is entering tech for the first time. i only bring that up because i have been shocked at how high the hurdle is for participating in saas conversations re early-stage startups. whether it's "figma for finance with a workflow element" or "carta for cap markets in the neobank space," the way silicon valley frames conversations around new software in terms of slightly less new offerings is astounding (it is both positive and negative. positive--faster to iterate on familiar concepts. negative--keeps other people out of the convo).
c) increasing distance between software greenfield and the average consumer. as i look at yc's large fintech companies, it seems like they've gone further up the value chain, or perhaps further way from the end-user. in fintech, you have stripe, which handles payments and which many companies might be familiar with (even though overall penetration has a lot of room for growth). but the "back-end" solutions like modern treasury or finch are not built for the average consumer, even if they can provide a ton of utility.