EDIT'ed in 'opening' remark:
What you are asking is in essence one of the central ethical questions: If you have to chose one life or another, how do you prioritize. It is a subject of much debate through the ages.
My personal stance on the subject, as you will see, is to apply the ethical framework that is most practical for the situation at hand. In this chase is a modified ulilterian approach (basically, which serves the most for largest quantety of moral patients under consideration). It's most practical-rating is established by the implied urgency of the situation (you can't take forever to decide) and due to fact that it is question that rates the survival of the moral patients higher than the following status of the moral agents (meaning, issues affecting those who act according to the system, like desensiblisation, are rendered mute by a higher value).
1.1 Suggesting five Phases
First you have to imagine that the evacuation of Capella probaly had multiple stages rather than being an uniform process. However this relies heavily on conjecture as to how the GTVA public was informed about the progress of the Nebula campaign; canon material only gives us the military briefings on the situation which vary in tone from full on probaganda (Return to Babel) and factual description of situation (Exodus).
When the official evacuation was first called the first Sathanas en route to Capella (and it was safe to assume that a shivan fleet was following it considering the Shivan activity in Nebula rose continously during the FS2 campaign) but the GTVA had a fair number of plans in motion to stop the Sathanas and they (reasonably) assumed that one of them would work in stopping the current threat posed by the Sathanas.
After the Sathanas' destruction the evacuation was not canceled as the Shivan threat persited and the best plan they had to end it had already failed (blowing up the Knossos and sealing the node to the Nebula); mark that down as the second phase.
A third phase could be set at the time when it was obvious to the military that the Sathanas was not a unique ship but that there was a fleet headed for GTVA space. A fourth and last phase can be assumed to have begun when the Sathanas fleet became common knowledge; though this one might not a phase on its if the information about the Shivan fleet coming to Capella was fully disclosed to the public at all times (whetever or not that is plusible is up to anyone; I personally doubt that the "progress" of the Nebula was at all times known to the public - preventing a panic among the people would be a good motive to hide that you are losing).
There could also have been a migration of certain groups in the Capellan population prior to the official evacuation. The threat of the Shivans might have been enough to already scare people into leaving before the GTVA saw it fit to take action. Take this as the zero'th phase of the evacuation.
1.2 Dynamics of each Phase
Having established that I would assume that dynamics of the evacuation would vary between these phases, including who ordered they were:
During the zero'th phase, it would have been people leaving who could safely leave their livelihood or their appointed positions. However their number might have been very low, depending on the number of avalible ships (which would have been the lowest of all phases due to the lack of government support) and range of people who actucally could leave without problems.
During the first phase, the GTVA may have encouraged people to stay on instead of rushing to the ships, saving up space in the ships to conform with the GTVA's priorities on who goes first and who goes last. Since it was reasonably safe to assume that the threat could be halted, there also would have been the assumption that the evacuees might return safely after the fighting was done.
Changing to the second phase, the GTVA may have dropped the pretence that it would be safe to return someday, but would have done so quitely as not to instigate a panic.
During the thrid phase, the incentives for the people to stay might have been dropped, it order to get everyone ready to leave but preventing widespread panic again.
Entering the fourth phase one can assume that the GTVA might be losing control on the selection of the evacuees here as there is no proactive stance against a panic is remaing. Rodo's approach might be in full effect here.
2. Applied Criteria during GTVA oversight
As for the priorities of the evacuation, they would only be in effect during the first, second and thrid phase, the zero'th phase being not a government encourage effort and the beginning of the fourth phase marking the loss of government oversight.
As for the pattern who to select to go first and who had to stay, I'd suggest a simple set of hard criteria:
a) Is the person essential to the operations of the military against the Shivans?
b) Is the person essential to the coordination of the evacuation effort?
c) Is the person essential to the logistics of the evacuation effort?
d) Has the person skills that can benefit any of the above?
Yes on any of these and you have to stay. No to all of these and you can go. Exceptions only, when the skill would be also put to good use at the destination of the evacuation, e.g. medical personell.
Than add a set of soft criteria which once you are on the "you go"-list prioritize your place there. While the hard criteria cater to the neccessities of the evacuation, the soft criteria are more open to ethical debate - going against the hard criteria means risking everyone, where as going against the soft criteria means upseting someone. The soft criteria also should be less hierarchical and be equally contributing factors to a weighted judgement. Some soft criteria might also be withdrawn if a certain quota was met (see. 3)
1) Age - the lower the better.
1a) Legal caretaker for (or if that is more your falvour:Parent of) a group or an underage child, already on a transport.
2) Health status.
3) Skills and aptitudes that serve at the evacuation destination.
1) and 1a) form the traditional "Women and children fist"-paradime which I think does not need an explaination.
2) is a matter to debate to some degree. Do we evacuate the sick and infirm first or last? Do we evacuate patients with contagious diseases last or not at all?
3) Is a matter to serve the needs at the destination, e.g. shipping out the doctors last might be detremental to evacuation effort as it would allow health problems to occur unchecked. One might object that such personell would be provided as part of the evacuation effort at the destination. This criteria might also be one subject to a quota of its own and eliminated as soon as the quota was met.
Adding other soft criteria might be appropriate depending on the additons to the ethical framework and its priorities that are applied, but IMO those three do cover it.
