Again, many of these things can be added later. Why such strong desire to complicate things? The OP is comparable to original embark - in fact, it has a higher degree of control (more complexity) already than original embark... It doesn't really need more, I don't think.
"But how do they make the journey safely?" How do your mountainhome expeditions do it safely?
"But you can escape from an ongoing demon attack." So what? You can also abandon and start a new game from the mountainhomes during an ongoing demon attack, or more directly relevant, you can look up legends, figure out when is the month OF a siege in the spring, and embark then from that civ and the settlers will get past a local likely sieging army somehow. Same difference. Already possible, and nobody seems to care.
"But how do they get to the edge of the map?" How does the original embark get to the edge of the mountainhome map?
"But you choose the dwarves and location in different order!" You don't have to - the idea would work perfectly well if the UI just let you pick a site then choose your list, if you prefer it that way for consistency. I don't care / have any good reason to have flipped it. Just an oversight.
"But squads only allow 10 people and that might need to get changed first." Except that the squad thing doesn't serve any need other than increasing coding complexity (you'd also have to recode bad drafting thoughts by the way). Just choose from a bare text list, what do squads contribute to the equation?
I'm all for adding in personally overseen adventurer wagon expeditions, and walking to the edge of the map, and things like that, as optional complexities eventually, but the majority of the benefit of the suggestion is in the raw basics, it's just as fair and challenging or not as original embark, and demanding half a dozen window dressings all at the same time that would take 3x as long seems misguided to me.