Gotta disagree on the differences in how easy it is to unionize or radicalize these different workplaces. It's not about a job simply being harder or easier, but about the forces that combat collective action, often intimidation but also mobility and power of capital vs. workers.

Example: organizing a restaurant can be difficult not because it's inherently harder for workers (front or back of house) to see class conflict, but because of the efficacy of the union busting playbook, intimidation of back of house that is often undocumented immigrants, and being forced to work directly alongside management 24/7. If that restaurant is a centrally-owned chain, even worse, as they'll just shut the whole thing down using some flimsy excuse. This can be harder to do for other industries that are less mobile and have fewer of these countervailing forces, e.g. longshoremen.

Example: unionizing Amazon is actually going very poorly. They have all of the hard labor things you might be thinking of and ALU sucks so they aren't getting anywhere. Amazon is, of course, in a very powerful monopoly position and gladly cycles through staff and does an effective job at union busting, but the issue here is poor organizing skills and a poor approach to the union overall. The Chris Smalls Show is fun to watch but terrible at actually doing the work.

In both cases, we need both things: a good understanding of workers' social relations to production (and therefore management) and competent and knowledgeable organizers that will run militant campaigns based on that understanding.

In my experience the vast majority of self-proclaimed socialists, including those interested in labor, don't have even basic organizing skills, let alone realistic strategy meetings or campaigns that are in any way serious. They have some enthusiasm and they like labor aesthetics. Some don't commit the effort or time, yes, but many actually throw themselves into it and get nothing done because they don't amplify their impact by doing the basics of organizing, choosing to take on activities that can be supported by 2 or 3 individuals that don't know what they're doing rather than 10 people ready to fight and distribute work. Or worse, just spend their time talking about what's happening and what take to publish about it in a newsletter.

I do see and know many folks like you're talking about. Borderline or actually burned out because when they try to organize, it ends up being a ton of work on them. While it's true that organizing does take a significant time commitment, it is also my experience that these are folks that failed to train anyone to be their comrades-in-arms, and thus still have some skills to learn. Or, very commonly, they're in an org that has a toxic culture where people don't really want to do organizing, they just want to play at it and pretend, so it feels like you've got no support.

When folks find an org where it's 100% experienced organizers, it's like a breath of fresh air where things happen rapidly and competently because they know the patterns and distribute the work. What I'm describing is basically the cadre model, though I usually avoid that term because it feels larpy.

The idea of labor aristocracy goes back to pre-Marx, where it was used to explain why the English trade unions failed to radicalize and could in fact be pretty reactionary and imperialist. The idea was that while they are proletarian in their relationship to production (the labor part), they receive so much gain from imperialism (along with social status) that they fail to become globally class conscious. Instead, they support their country's imperialist wars that bring home the loot they split with their bourgeoisie.

This is reminiscent of US labor that was "bought off" by similar means a century ago and many of the remaining industrial unions in the US. Those for military contractors are the most obvious, but a lack of global class consciousness can be found in almost every union here. Knee-jerk support for other unions kicks in, so you'll see Teamsters and SEIU and UAW supporting Boeing workers next year despite the latter making the tech that eventually spies on Gazans and bombs kids. And in only one of the above is there a subculture that I'd call class consciousness (UAW via UAWD). Try having conversations about socialism in these unions and you'll have to scrounge for even basic class consciousness and might get hounded out by staff. The classism I mentioned earlier is rife in two of those big unions.

This is something we have to recognize if we want to rely on a labor strategy as socialists in the imperial core. I've seen a lot of socialists with very little labor experience but rose-tinted glasses about labor militancy from the late 1800s run head-first into union organizing and then becoming dismayed at what they find when they stop looking for the "corrupt union bosses" as the only ill. And those socialists that are anti-imperialist will quickly find that labor-focused socialists tend to be imperialist, as their efforts do tend to be in support of the conditions of a better-off subset of the imperialist working class. Try telling them to have a slightly critical approach to the Teamsters, who are both militant and reactionary. They'll start to get scared they'll lose what little labor connections they have (not a false fear).

Anyways, that's the old-school concept of labor aristocracy and also more or less the same concept used by MLs and Maoists when they want to look at this problem. It's an attempt to provide a material interest rationale for why the imperial core trade unionists almost always suck at class consciousness, especially anti-imperialism.