That geographic proximity makes other differences. The situation the players want to engage is (in) City B. The siege is a potential resource for the players to leverage in respect of doing stuff in City B. To use a semi-technical phase, the
fictional positioning is completely different in the two cases:
* desert: until the GM introduces more content (per Hussar, "lays a trail of breadcrumbs"), the fictional positioning of the PCs does not give the players any capacity to leverage the desert in pursuit of their goal;
* siege: the fictional positioning of the PCs gives the players capacity to leverage the siege in pursuit of their goal, to be proactive and require the GM to respond to their plans (eg sneaking in under cover of bombardment).
So the fact that the siege IS a breadcrumb, and the desert would need to have a breadcrumb ADDED is now relevant? When the Players say "we go through the desert to the city", their goal is, at a minimum, to enter the city. If the nomads, or the refugees, or the siege, get in their way, the result is the same - the players are blocked from their goal of entering the city by something the GM placed in their way, which they did not know about until they came across something which made them aware of it. They might be able to use the nomads, the refugees, or the siege, to advance their goals. Certainly, the fact they are in the way affects their goal. But the siege can be one which has no relevance, and the nomads, refugees, or anything else in the desert, can be relevant.
There is another dimension, of course. They can each lie in a different area between "highly engaging" and "extremely boring". But that can't be known in advance, and can be the case regardless of relevance.
Until the GM runs some encounter or other, the players have no resources to leverage in pursuit of their goal. Whereas the siege is, per se, a resource that the players can leverage.
Until the GM runs the encounter with the siege (even so simple as "off in the distance, you see tents, wagons and creatures/people milling about), players have no resource from the siege. Like the nomads, or the refugees, they have to encounter it first.
The nomads are not inherent in the desert. The players have to narrate their PCs entering the desert and wait for the GM to frame a nomad encounter (in Hussar's terms, wait for the GM to lay a trail of breadcrumbs). This is reactive, not proactive.
The players have to narrate their PC's crossing the desert ("we ride our centipede") and wait for the GM to frame a siege encounter (in Hussar's terms, wait for the GM to lay a trail of breadcrumbs). This is reactive, not proactive.
Whereas, with the siege, the players aren't waiting for the GM. There is something to which they can react.
Once the GM frames a siege or nomad encounter, the players aren't waiting for the GM. There is something to which they can react. Until the GM frames the encounter, the players are waiting for the GM.
But, that presupposes a certain playstyle where the physical reality of the game world is such that only the DM can determine when parts can be hand waved. It's very much like an MMO game world. If I want to travel in an MMO, I have to travel through each of the regions in between, risking being attacked or whatnot on my way from start to finish. ((Note, I'm not trying to bring in a video-gamey canard here, just making an example))
The major difference I perceive from the MMO world is that the MMO world requires you to kep pushing the "go forward" key to cross the desert. It does not fast forward until something (whether somehing you want or not, whether something you expect or not) interrupts your journey. A a minimum, in game, the "step forward" time should not be played out. I've never seen a GM require the players to state each step "left foot, right foot", until arriving somewhere.
I don't, and I get the feeling that Pemerton doesn't, feel that the DM should have the sole say in what parts of the game can be hand waved. Yes, it would likely be best had we had the proper in-game resources, but, we didn't. But, since we're playing a game, I don't feel that the players need to be forced to play out the consequences, when they don't want to play out those consequences. Most of the time, players will want to play out the consequences. Your ship sinks and you make it to the desert island - presuming buy in from the players, you are good to go.
But, I also don't have a problem with the players having strong enough opinions to say, "Hey, y'know what? Let's skip over the Castaway scenes and I don't even have a volleyball - can we just get to the part where we leave the island?" Obviously, some people do.
Again, this comes back to two issues. How frequent (and I think you have answered that as "very infrequent"), and what happens if the players lack the hive mind to all agree on which scenes shoul be played out? Whoever is the easiest bored and most finicky about the scenes played calls the shots? All players must agree? X% of the players must agree? Here, I generally defer to the GM, but would have no problem with the GROUP perceiving things dragging in an area suggesting we cut that aspect short. The group. Not any one single player.
On a side note about relevance between the desert and the siege. Imagine for a second that the party does have a teleport spell. Now, they've arrived in the city with the city under siege. The desert in this scenerio is entirely irrelevant. Is the siege still irrelevant? I would say no. If nothing else, the players must achieve their goals before the siege breaks into the city. Plus, the presence of the siege is going to radically change the reactions of the NPC's. A peaceful city is going to be completely different than one under siege. However, the presence of the desert around the city doesn't really have much of an effect - the city happens to be in a desert and that's about it. It's entirely background and has very little impact on play.
It will or will not have an impact as th game plays out. If the siege stays outside and the PC's stay inside, the siege is entirely background and has no more impact on play than the desert outside.
That's why the siege is relevant. It's always relevant no matter what the players do. Even if they just fly over the siege, the city is still under siege. But, the desert? Doesn't really matter a whole lot. The nomads are still out there somewhere but, since they aren't linked to our goals, who cares?
I suppose we could "randomly" meet that group of nomads on our way to the city, who just happened to be the nomads that we need to meet, but, that's about as contrived as you can possibly make it. Our random Planar Travel destination results in our landing in exactly the right place to meet the nomads? If this was a novel or a movie, I'd stop right about there. That's Syfy level writing.
And yet this is about the only way Pemerton finds the nomads acceptable at all. Different playstyles indeed.
Yes, that's certainly a playstyle thing. I think the relevant phrase I've heard bandied about in the past is "meaningful choices". There has been around here a sizable chunk of people who want/need the game to be largely about player choices, and the consequences of those choices. If you get to say, "I just don't feel like playing through the consequences," then you remove the meaning of the choices that led up to that situation.
Now, if you were on a railroad prior to that, this argument doesn't really matter, as you didn't have many meaningful choices up until that point. But, as broad generalization, a case of, "I don't feel like it, please fast forward" will tend to run up against this issue.
An excellent point. You made the choice to recruit mercenaries, you make the choice how much or how little work to put into the vetting process, you play out the consequences.