For what it's worth they're the only ones really having a go at the ICANN DNS root. Most other projects are nibbling around the edges, or just pretending.
About no HTTPS, it only makes sense. Certificate authorities are built to query the ICANN DNS system for domain validation, and HNS is off the ICANN radar, ergo invisible to certificate authorities. But then there's DANE, which is arguably more secure than HTTPS anyway, and relies on DNSSEC to secure DNS records, and HNS domains can support DNSSEC, so it's not a question of HNS being inherently insecure. You can see having to resolve DANE (or similar) as just part of the wider resolution paradigm.
The point of HNS domains is it's a decentralised, permissionless alternative to the entire ICANN-managed DNS root. Basically have your domain controlled by a keypair, which is kind of like Nostr identity for domains in a way.
Why a blockchain versus an auth database is meat and potatoes decentralisation. If you're going to have a go at ICANN you won't get very far with an auth database. You could argue the tokenomics is a little messy and HNS should just go full bitcoin (boring old proof of work, fixed supply, etc.) but however you slice it you need some such foundation.
There's also PKDNS which could resolve to PKARR using some kind of human-readable woo woo, but that's a whole other kettle of fish.