I think that due to how much F9 has managed to eat into FH's market, FH flights we see are more likely to only happen when the payload is heavy enough to need the center core to be expended.
For example, for a GTO insertion, an expended F9 can do 8.3t while a completely reusable FH can do 8t, but the expended F9 is going to be cheaper overall. Thus FH is only worth using when the payload is heavy enough to require at least the center core to be spent. This might be different for Lunar payloads though.
When I skipped through the video to the point where they explain that this was a mission without center reuse I was surprised to hear that they actually do seem to offer a three booster reuse profile. Perhaps my surprise was from confusing "no second stage reuse" with "no center booster reuse"?
But if it's a wash (as you say) between non-reusable F9 and 3x reusable FH I'd probably go for a single EOL core anyways, for the lower total count of failure modes alone. Even more so if launches in that weight class are sufficiently rare to be served from "organic" churn of reusable F9 launches (they clearly are).
In any case, my mention of surprise in the top level comment was not meant as "how wasteful, losers, boooh!", but as an statement of amazement, "have we/they really come that far that a part of me is unimpressed by two of three?", and on a meta level surprise about how a non-reuseable F9 feels less disappointing to me: a non-reuse F9 is a valid compromise to eke out the last bit of payload, whereas the FH raises an expectation of being better in every way that's then disappointed by "a third of a step back" in terms of reuse compared to what we now consider a regular F9 launch. We consider the reusable F9 regular, crazy!
Do you have any source for this? I would expect that a FH launch that reuses all 3 first stage cores would be cheaper than an expended F9. A few hundred thousand extra in fuel costs for the side boosters is cheaper than the few million dollars that the 9 expended engines would cost.
But I still believe you -- the infrequency of an FH launch probably increases the ancillary costs. I'd just like to see your source.
As for the price, the official numbers are $67m for a new F9 and $97m for a reusable FH (from their respective Wikipedia pages).
Additionally, my reasoning is (as you mention) related to cadence. FH flies very infrequently and requires more ground expenses to prepare, recover and then refurbish 3 boosters. Additionally there's the opportunity cost of keeping a center core on standby (as it's purpose built for FH [1]) and the opportunity cost of keeping the side boosters either on standby or converting them to F9s.
Elon has previously mentioned that with Falcon rockets the main costs per flight are the second stage (~$10m, expended), fairings (~$3-4m, reusable), refurbishment (~$1m) and fixed ground costs, which can be reduced by increasing cadence [2]. In the case of FH the second stage and fairing costs are the same, the refurb costs are tripled and due to much lower cadence the ground costs are higher. Meanwhile, F9 flies frequently, so it has lower ground costs per launch and the booster's cost can be more easily amortized via several flights before the expendable launch.
Additionally, [2] mentions that the marginal cost of building an F9 booster is ~$15m, so all the extra costs of FH only need to be greater than $15m to make it worth just using an expendable F9.
Hmm yeah, digging around it seems that Wikipedia is probably incorrect with the expendable pricing. Seems like there aren't really any recent numbers available for expendable F9s as the last intentionally expended F9 was 3 years ago.
For example, for a GTO insertion, an expended F9 can do 8.3t while a completely reusable FH can do 8t, but the expended F9 is going to be cheaper overall. Thus FH is only worth using when the payload is heavy enough to require at least the center core to be spent. This might be different for Lunar payloads though.