The gemara is trying to figure out the status of a chatas that belongs to the tzibur when one or all of the members dies, whether it has a status of חטאת שמתו בעליה that must be killed. The gemara tries to prove that this concept doesn't apply to a korban of the public from goat that is sacrificed on Rosh Chodesh as a korban chatas from the money that was collected at the beginning of the year. The fact that we aren't concerned that one of the contributors died thereby rendering the korban invalid, indicates that we don't apply this concept to a communal korban. The gemara rejects this proof because by the Rosh Chodesh korban we aren't sure that someone died, but in the case we are dealing with (sacrificing a chatas for the generation of tzidkiyahu) we know that they have died. Meaning, that when we are in doubt we can bring the korban, but when we know for sure that a partner in this korban died, we can't bring it.
The Mi'lo HaRo'im asks, if when we would know that someone died it would invalidate the korban, now that there is a safeik that perhaps someone died, the korban should also not be brought because it is a safeik of bringing and invalid korban into the beis hamidash - why don't we invalidate the korban of Rosh Chodesh based on safeik d'oraysa l'chumra? The Mi'lo HaRoim answers this question by saying that we can use a chazaka that the contributors are still alive. But, he immediately asks that the chazaka only applies when there is a possibility that the contributors are alive, but by the time you are most of the way through the year, there have DEFINITELY been people who died?
Tosafos HaRosh asks this very question. Every single Jew contributed toward the korban, by the time the month of Adar rolls around there is no question that one of the contributors had died - how can this korban be brought? Tosafos HaRosh answers that although some people have definitely died, we aren't sure that this money that is being used for the rosh chodesh korban is from one of the contributors who died. Since there is a possibility that the money we are using for this korban is from people who are still alive, the korban can be brought. Why can we assume that? Because there is a chazaka that the contributor of this money that is being used is still alive, therefore using the chazaka that all contributors to this korban are still alive, the korban can be sacrificed.
Another possible approach that the Mi'lo HaRo'im rejects is that according to the Rambam that safeik is l'kula on a Torah level, one can be lenient that the contributors are still alive and sacrifice the korban. He rejects this because the Rabbonon are stringent about a safeik d'oraysa so they shouldn't have allowed the korban to be brought. The Mitzpeh Eisan on the other hand does consider this a valid approach. Perhaps the Rabbonon weren't machmir in this situation since it would prevent the ability to sacrifice the goat of Rosh Chodesh. He suggests that even according to those who are machmir for a safeik d'oraysa, that would only be for a d'oraysa prohibition that is explicit in the Torah but not for something like חטאת שמתו בעליה which is just a halacha l'moshe misinai to invalidate the korban - all would agree that safeik on halacha l'moshe mi'sinai is l'kula.