Honestly, it's actually a little surprising that they're not cheaper than everyone else. It's direct marketing only (no commissions), underwriting requires no staff, one product, etc - all of that likely means much lower expenses, which is the holy grail in term pricing.
As far as the actual underwriting, the only piece they're really missing is the added info of the medical exam (primarily the blood work). It's not so crazy to think that in the presence of all the other hard data they might be able to pull, the medical exam doesn't often meaningfully affect the outcome. Layer on that any conservatism they likely built into their mortality assumptions for lack of prior experience with the system and it doesn't seem like that big a risk.
The technology costs money - less than an agent once you reach scale, but they're not there yet
Advertising costs often eat up as much as commissions or more
The underwriting costs money - someone has to pay for the R&D to develop those models, plus it costs money to ping PBM, MVR, and the other data sources (which I'm not allowed to list). Plus, of course there's some underwriting staff - they need people to review after the fact to ensure model quality and people to handle applicants who don't qualify instantly.
Term isn't very profitable in the first place, and if you don't have scale it can be very unprofitable. Especially for incumbent companies who have high overhead cost or have high hurdle rates to justify branching out of their core business. Ladder or Quilt may have better ability to undercut the market - though their current prices show no indication of doing that.