First of all, for the record: The most basic harmony is as simple as laying Luke 3-9 next to Mark 1-6 and weaving both into one, without alteration of either sequence. Also note: The simplistic nature of that exercise does not automatically invalidate its historical potential. Notoriously, Gospel Harmonization is as difficult to deny as it is to confirm.
If we could be assured that both writers are giving us perfect sequentiality, the only logical conclusion would be that, yes, similar things must have happened twice, on at least two occasions. Lacking that assurance, conservative seminarians have embraced "theological redaction" as a means of explaining these narrative differences. I suspect they mainly find it more defensible. I myself find it less historically sensical.
For one thing, if Luke altered Mark's sequence to emphasize different thematic aspects of those stories, then why did Luke not do so more often? More substantially, if Luke's differing details are based in fact due to some alternate source(s) other than Mark, then how do we know the alternate source(s) didn't describe at least one alternate event?
Some evangelical scholars dismiss sequentiality in Luke's Gospel simply for lack of "chronological" language. They point out that transition phrases between episodes in Luke's Gospel aren't necessarily explicit about the historicity of the narrative sequence. Language, phrases, narrative: christian exegetes seem to have a verbal fixation. Sometimes I'd swear they think events are just vehicles for relating ideas. (I'd say 'think about that at your next protestant church function' - but then I'd be digressing.)
Okay, now. So what else can be done?
One major purpose of studying History is to note the significance of dynamic events. Julius Caesar crossed the Rubicon many times, but that one famous time he did so, it changed everything. Therefore, one way to analyze the Gospels' event sequence is to look for dynamic events. Which moments altered the status quo? What do Mark and Luke have to say about that? About 'game changing' aspects of certain persons' behavior? About causal relationships between actions that take place in narrative?
Here's what I am suggesting: If the writer relates an event in such a way that it would make little or no sense (as written) being moved to another stage in the 'plot', then the writer has explicitly implied the sequentiality of that event. In other words, if the narrative presents any episode as necessarily pivotal, and the surrounding narrative illustrates a permanent alteration of the status quo (the 'narrative sitz im leben' we might say) both before and after that episode, then that historical event has been deliberately located within the writer's event sequence, and placed there with chronological intent.
Furthermore, if two writers seem to hold different opinions on the dynamic significance of an event, our analysis of their accounting for that event could hold keys to determining historicity (and chronologicity) of the historical event sequence at large, or at least portions of it.
Obviously, I'm still working out these ideas. For a first illustration of what the method might look like, come back on Monday and think with me about the fishermen calling(s?).