Ok, let's start with the regs:
14 CFR, part 25.721
(b) Each airplane that has a passenger seating configuration excluding pilots seats, of 10 seats or more must be designed so that with the airplane under control it can be landed on a paved runway with any one or more landing gear legs not extended without sustaining a structural component failure that is likely to cause the spillage of enough fuel to constitute a fire hazard.
14 CFR, part 25.562
Emergency landing dynamic conditions.
(b) Each seat type design approved for crew or passenger occupancy during takeoff and landing must successfully complete dynamic tests or be demonstrated by rational analysis based on dynamic tests of a similar type seat, in accordance with each of the following emergency landing conditions. The tests must be conducted with an occupant simulated by a 170-pound anthropomorphic test dummy, as defined by 49 CFR Part 572, Subpart B, or its equivalent, sitting in the normal upright position.
(1) A change in downward vertical velocity (Δ v) of not less than 35 feet per second, with the airplane's longitudinal axis canted downward 30 degrees with respect to the horizontal plane and with the wings level. Peak floor deceleration must occur in not more than 0.08 seconds after impact and must reach a minimum of 14g.
: 35 feet per second is 2100 FPM, max vertical speed at landing that the plane is required to withstand before the gear fails is is 600 FPM, so this provision covers both a gear-up landing and a hard-beyond-design-loads gear-down landing, any "normal-like landing" but with one or more gear up, must be much more smooth than this, not to mention if the pilot manages to make a smooth landing.]
(c) The following performance measures must not be exceeded during the dynamic tests conducted in accordance with paragraph (b) of this section:
: what follows basically means that you are uninjured enough to easily evacuate the plane on your own]
(2) The maximum compressive load measured between the pelvis and the lumbar column of the anthropomorphic dummy must not exceed 1,500 pounds.
(4) The lap safety belt must remain on the occupant's pelvis during the impact.
(5) Each occupant must be protected from serious head injury under the conditions prescribed in paragraph (b) of this section. Where head contact with seats or other structure can occur, protection must be provided so that the head impact does not exceed a Head Injury Criterion (HIC) of 1,000 units.
(6) Where leg injuries may result from contact with seats or other structure, protection must be provided to prevent axially compressive loads exceeding 2,250 pounds in each femur.
(7) The seat must remain attached at all points of attachment, although the structure may have yielded.
(8 ) Seats must not yield under the tests specified in paragraphs (b)(1) and (b)(2) of this section to the extent they would impede rapid evacuation of the airplane occupants.
So, as you see, the airplane is required by regulations that a one-or-more gears-up landing don't cause a fire, and that an impact much stronger than a gear-up landing don't injure you to the point where your evacuation would be compromised.
In other words, the airplane may be a write-off after a gear-up landing, but passengers must be safe.
Now, previous cases.
There've been several full and partial gear-up landings. Maybe a couple of dozens. I agree that's not a very big sample size, but the fact that the passengers seem to always walk away uninjured is comforting, to me at least.
Interestingly enough, the vast majority of the few total gear-up landings was the result of pilot error (pilot not extending the gear and not performing the checklists, sometimes coupled with the failure of the gear warning, sometimes the warning got silenced by the crew, and sometimes the landing config was one that would not trigger the gear warning).
I say "interesting enough" because in these cases, where the pilot was not aware of the condition, he could have not planned and executed a particularly smooth gear-up landing. For the pilot, it was a normal landing until the "grinding noise" started. But still, in each case everybody walked away on its own. You'd hardly commend the pilot for that. The good outcome was "despite" the pilot actions, rather than "thanks" to the pilot actions.
The LOT case was partially an exception. I mean, they didn't forget to lower the gear or to perform checklists and they performed all the abnormal/emergency procedures and checklists. But a pilot that knew the systems well it could have thought of the alternate-gear-extension motor CB, or any pilot could have checked all the CBs to see if there was something odd. So the human factor in the cockpit is not absent here.
I claim that procedures and checklists MUST work without the need of pilot creativity or proactivity. But if it doesn't, then pilot creativity and proactivity is welcome. Not that he would do anything on his own (he could be unintentionally worsening the situation), but he can consult with the engineering and maint teams on the ground. That's CRM. And in this case, it would have prevented a gear-up landing.
I've seen a few gear-up landing procedures, and at least those few didn't make a difference between a one-gear-up, multiple-gear-up, or all-gear-up landing. All of them called to land on all and any landing gear that was available and to make a normal landing, except for some configurations (like flaps, spoilers, autobrakes, etc.). There seems to be an agreement that it's better to land on any gear that is available than with all the gears up. The rationale is that the most energy that can be absorbed by any gear, and the less grinding of airplane parts on the tarmac, the better. Controllability doesn't seem to be a major problem, except when already at low speeds.
Check these videos for a landing on one main and nose gear, which seems to be the condition that worries you most:
the comment was that full landing gear failure is virtually unheard of and never trained for.
Don't believe everything you see on TV. No, wait. Don't believe ANYTHING you see on TV
A full gear-up landing, while thankfully very rare, is not unheard of. And the procedure, and hence the training, seems to be the same for an all-gear-up landing or a partial gear-up landing. I think you'll like this video:
A part of your comment (now deleted) that upset me (not your fault) was that there was no info from the flight deck about what was going on and what to expect except for the "brace brace brace" commands (or something like that).
I think that doesn't speak well of this captain. I think that something like this would have been in order:
"We have a landing gear malfunction that we could not troubleshoot, so we will make a belly landing. I want you to know that the airplane is designed for such an event and that we are trained and have procedures to deal with it. We'll have the runway foamed and we'll make a smooth landing on the belly, and the plane will skid to a stop. As a matter of precaution, the airport's emergency crews will be waiting for us, we'll assume the brace position just before landing (I'll tell you when) and, once the plane has fully stopped and on command of cabin crew, we'll evacuate the airplane as swiftly and orderly as possible to avoid that any of you gets injured in the process. The cabin crew will give you further information regarding the preparations for the emergency landing, the brace position and evacuation. Always follow their instructions."
I'm also curious to know if these CBs are part of any checklist (maintenance, pre-flight, etc)? Is there any possibility that the CB had been open for a while, but only became an issue once the hydraulic failure occurred?
That's a good question and, together with protecting the CBs from unintentional manipulation and including the check of the CBs if the alternate gear extension fails, seem to be at the core of the incident.
Have you read the report, especially page 10?
I don't know in the 777, but in the checklist of the Tomahawk I used to fly there was a "CBs --> all in" line.