It seems like the slow speeds of airships will only be useful in niche applications.

True, long-haul flights will have an overwhelming time advantage. However, those are only about 5% of all flights. Over 80% are short-haul flights where airships would excel both in terms of efficiency and operating costs, with little loss in relative time due to the lower distances involved.

Airships aren’t actually that slow. Over short distances of a few hundred miles and carrying 100 tons of payload, the most productive cruising speed for a hybrid, heavier-than-air airship is 180 knots, and for a conventional, lighter-than-air rigid airship, 145 knots. Both are a fair bit faster than the average helicopter (~120 knots), and on the upper end of bullet train top speeds (and given how often trains stop and decelerate, their actual average speed is far lower).

However, longer distances demand slower cruising speeds. The optimum for distances of a few thousand miles is 150 knots and 82 knots, respectively. Hence why I was looking at 100 mph cruising speeds for a midsized conventional rigid airship, which converts to about 87 knots.

Maybe international cargo to split the difference in time vs costs of ocean vessels and standard aircraft.

If you want to keep your airship below the length of a cruise ship or fuel tanker, your upper limit is about 500-1,000 tons of payload, depending on the design. That won’t replace bulk ore, fuel tankers, or anything really heavy and cheap, but it is far cheaper than conventional air freight per ton-mile, which is good for things like high-value manufactured products and food.

A significant additional benefit is the avoidance of transshipment costs and delays. Airships can land in any open field with enough space, or even lower down payloads while in flight with a gantry that just needs to be filled with water ballast or additional cargo while offloading, thus don’t need airports, warehousing, rail transfers, trucking, etc.

I can imagine a future where most international shipping goes by airship, and mostly in the direction of the trade winds. That doesn't seem likely, but possible.

Probably not most shipping, but a really good chunk of it. Particularly for bulky cargoes like aircraft and rocket parts, wind turbine blades, mining equipment, and so on that can’t physically fit inside most airplanes.

Maybe very short passenger trips like San Francisco / LA. And only then when the weather is suitable.

Short ferry flights are the most promising passenger application, yes. The weather is less of a big deal than you might think—most airships today are only small vessels used for advertising and sightseeing, but even the Zeppelin NTs used by Goodyear have a wind limit for landing and takeoff of 35 knots—which is the same as the limits for a modern airliner like the Boeing 737. The all-weather Navy radar blimps used in the Cold War had an 88% mission availability during inclement weather like blizzards and thunderstorms, which is better than most military aircraft can manage even in clear air today, and their wind limit for takeoff and landing was about 45 knots—which is towards the upper extreme of a literal gale-force wind.

As a general rule of thumb, the faster an airship, the higher its wind limit. Those Navy airships had a top speed of 82 knots, and a modern hybrid could be designed to be much faster than that.