Generally speaking, it is near-impossible to be 100% sure that your interpretation of a historical source is the "right" one. No matter how hard you work to make your analysis as correct as possible, there will always be at least one historian that completely disagrees with your analysis. For very recent textual materials (say, from the last 60-80 years) it is much easier to get the interpretation right, because we still have access to a lot of contextual data, and in some cases even the person who wrote the materials might still be alive. But the further back you go in time, it becomes increasingly difficult to interpret the meaning of the words accurately. For very old texts in the medieval or ancient periods, a single misplaced punctuation mark or a single misinterpreted word can completely change the meaning of a sentence or even a whole paragraph.
It just can't be helped. In the same way as scientific research, the field of history is one where each new generation tries to build up upon the work of previous generations. We try to pick up the things that the previous generation got right, and by accessing new materials and new research techniques, we try to correct the misinterpretations and errors found in their papers.
What is really important is that you conduct yourself with absolute professionalism at all times. Even if you find a blatant error in a history paper, always keep in mind that that person(s) was trying to do the best work they could with the resources that they had available at the time. The same is valid for your own historian colleagues still producing papers right now. Each historian has their own set of specialized skills, but they also have aspects in which they are deficient (ex: insufficient foreign language skills), which can lead to misinterpretation of historical texts. In sum, my point is: even if you completely disagree with the methods/interpretive analysis of a colleague, always explain your disagreement with sincere kindness and respect. In the future, if someone finds an error in your work, you would like them to treat you in the same way.
Regarding Question 1, at the very least you have the duty to get a basic idea of what the marginalia is trying to communicate. Even if you don't understand all of the scribbles, try to get a basic grasp of any passages that you can understand and search for other papers that analyze the same materials, to see what they have to say about the marginalia. If you think they are from a much later time period and are thus unrelated to your research topic, then you can safely ignore them. Don't be so quick to assume that these materials have not been studied much before. It is often the case that foreign historians have analyzed the materials and published papers about them in foreign languages. If you can't find any paper that deals with the topic, try to consult with a colleague or a supervisor, or try to contact by email a historian who has analyzed these materials in the past. In that case, when you write a paper, just be honest and mention clearly that you did not take the marginalia into consideration when analyzing the materials, and briefly explain why.
Regarding Question 2, your duty as a historian is to make the process of your interpretation clear to the readers of your paper. Whenever possible, try to provide an excerpt of the original passage in your paper (as a footnote), be brave enough to present your own English translation of the passage, and then explain why you decided on that particular interpretation of the text. Obviously you cannot do this for every single textual passage, but at least you should do this for very important passages that have a direct bearing on the topic of your paper. The reader should be able to follow your train of thought, and then decide whether to accept your interpretation or reject it or correct it. The value of your paper does not necessarily lie in providing the reader with the absolute "right" interpretation of the text: it is about being transparent about the methodological choices that you made when interpreting the text. In this manner, you are providing a service to the reader, you are providing clarity to them, and you are saving them time because they can evaluate your interpretation much more quickly.
Of course, you are still wondering how you can improve the accuracy of your interpretation. The secret is to collect as much information as possible about the context and conditions in which the text was written. The more contextual information you have surrounding the text's production, the easier it will be for you to discard certain interpretations in favor of others. It will also make it easier for you to interpret ambiguous passages. Unfortunately, many historians merely stick to investigating the social and economic conditions surrounding the text's author or the local community. Another potentially problematic "quick fix" solution is to merely apply some general theoretical interpretive framework on top of the text (ex: deconstruction, literary criticism, etc.), without taking into consideration other methods. People are busy, have families to take care of, and other responsibilities, so this situation is somewhat understandable. But if you can, please try to gather contextual information from other fields such as archaeology, environmental history, even paleogenetics. These fields can uncover traces of environmental disasters, epidemics and parasites, economic trade, bad crops, famines, daily habits, mass migrations and burials which can completely change your initial interpretation of the actual text. You can sometimes realize that certain ambiguous or obscure passages of the text are actually referencing times of turmoil and social anxiety caused by various environmental phenomena.
Another thing that you should try to do is take a general look at other prime materials produced around the same time around the same geographic region, and try to get an idea of how the history of that place was before the text was produced, and how it became after the text was produced. This will help you to situate the text in the broader historical conditions/trends of that region. Also try to find texts by other contemporary authors who show a different perspective around the same historical events. Identify which parts agree and disagree, and make them clear to the readers. By multiplying the number of subjective interpretive frameworks, and by multiplying the number of methodological analyses of the same text, you can actually paradoxically get a more objective interpretation, instead of a more subjective one. The more you remain in a methodological echo chamber, without considering competing perspectives, the more likely you are to get something wrong. Of course, you cannot realistically try every single approach. Just do the best you can with the time and resources that you have at your disposal.
Finally, if you are wavering between two or three possible interpretations of a textual passage, explain to the reader the merits and demerits of each one, and then be brave enough to choose one, with courage and dignity, even knowing that in the future you might be proven wrong. By presenting your work with humility, and by identifying the strong points and weak points of your interpretation (i.e., the parts where your hypothesis might be wrong), you will still be respected by your peers, even when you get some things wrong. And make no mistake, you will get some things wrong for sure! Be at peace with that fact.