AA can't fly as many pax back to the mainland as they fly from the mainland to Hawaii
 

I just read another article worth sharing from View from the Wing.

Gary quotes someone from AA saying:
I don't understand this--is this because of cargo demand from Hawaii to the continental US? I would think there would be more demand in the other direction for cargo.

LIH-LAX is 2,615 miles. The runways at LIH are just 6,500 feet long.

Here's my understanding of it:

Shorter runway => requires reduced takeoff weight => aircraft must carry less fuel OR pax OR cargo

Carrying less fuel is not a good idea, given that there are no intermediate/emergency landing points between Hawaii and the mainland.

So, the only alternative is to carry fewer pax or less cargo.

Prevailing winds (including velocity and direction), runway length and those winds, temperature, altitude and pressure (density altitude) and aircraft capabilities and restrictions (weight, including cargo, passengers, crew and fuel, runway length required for safe takeoff, crosswind component, etc.) and could all play a role in making those determinations. They’d be applied flight to flight, at or near time of departure.

Sounds like a job for the new 757! ...er, wait.

I suppose they could also add a tech stop at HNL to refuel and then depart for the mainland, similar to the way the La Paz -> Santa Cruz -> Miami flights operated.

I don't believe La Paz was a Tech Stop.

No; Santa Cruz was a tech stop. The origin, La Paz (El Alto Airport), is so high (13,325 ft) that even the 757 cannot take off with enough fuel to make it back to MIA. So, the bird took off with a minimal fuel load, then made a stop at the much-lower Santa Cruz airport to take on enough fuel for the intercontinental flight.

Veering OT, but if VVI was indeed originally intended as a "tech stop", it seems to have surpassed demand expectations, as AA has dropped the LPB stop altogether and now flies MIA-VVI as an out-and-back.

Anyways, as I've posted elsewhere, that short runway at LIH might also have contributed to the loss of DFW-LIH this year. That route was operated by the 767 last winter. However, DFW-KOA going to a 777 this year indicates to me that AA wants to consolidate all DFW-Hawaii flying onto the 777, which cannot make it off the LIH runway without taking a significant payload hit. As a result, the route got canned. Sad!

My last 2 flights out of LIH have had to kick people off the plane to allow it to fly. The last one was done after everyone had boarded and we were ready to go. Offer started at $1000 and they needed 5 people. I think it got up to $1200 for the last 2 people. That's a lot of money to throw away because you don't have the right plane (or planning, really) to operate the route. Almost took the voucher but First was sold out for the next two days and I needed to be home the next day. This brings up the other issue - there are not enough premium seats on these flights any more.

Or when AA began MD A Lemon, er MD-11 Flights SJC-NRT, requiring a Technical stop for full fueling at OAK.

AA is challenged to come up with exact fit” aircraft, even with the A321neo. (The 757 is an aviation “hot rod” for hot, high and shorter runway lengths.)

Well, AA could have kept enough 757s for Hawaii service, until the 321neo/737MAX8 were certified, but someone (DP) didn't think far enough ahead. Very short-sighted management decisions.

The other issue is why are they waiting until all passengers are on the plane. That is simply poor operational planning. If they need to block seats regularly, why don't they simply reduce the capacity of the plane for booking purposes???

In the context of buying a bigger plane to serve that one route with occasional weight restrictions, $6K isn't really that much money.

List price on a 321 is $115M. That's $6K daily for 52 years.

I wonder how a 757 Max would be selling right about now. Too bad, Boeing didn't try it...

Keeping a subfleet that's less fuel efficient for a particular route doesn't exactly sound like it would be a "good" management decision. I'm not sure why you believe this is short-sighted - the choice could well be current aircraft, or none at all.
And certainly the other issue cannot be planned for in advance. Winds, routing, the amount of baggage each person brings with them, not to mention no show patterns are all variable.
Sure, you could always undersell the flight. But then you would be preventing people from purchasing tickets and getting home on the day that they want when on certain days, the aircraft could have made it just fine. And underselling would eat into the financials of the route, which could mean that not operating it at all becomes more economical.