A small disclaimer: I'm coming from an medieval / early-modern background, and this answer isn't completely applicable to more history.

I think we may need to come together and work out a little of what it means to "know" in history. I know that sounds really pompous, but I think it is pertinent to this question because you used the word reliable, which I am taking as a synonym for accurate. So my answer is going to be on historical method, and not on the Goths--sorry about that.

Historian's reconstructions of societies are not 100% accurate. Period. Think about what the historian does, which is compile sources into a narrative, and then think what goes into that. We can take the Goths as an example. It all starts with sources, or at least what's left of them, because the society is hundreds of years old. We cannot read them in their original context, so much will be lost on us. Some of these sources will be primary materials, others will be from other distinct geographic groups. Much of the time we do not know that much about the authors, which makes it all the more difficult to pick out textual biases. All of these issues (and more) are within the sources themselves; we haven't yet gotten to the historian themselves and their role, which I don't want to elaborate on here.

I hope that the above glimpse of the problems of sources can help to show you that any work of historicism is a messy patchwork of information, which can never properly "represent" its subject. That is, if we could magically time travel to a historic subject, it would not totally resemble our historical portrait. Thus, to go back to your question, most knowledge about gothic history cannot, and never will be, considered totally "reliable." The same is true for all historical periods, but especially so for those that didn't leave behind massive paper trails.

(Sorry for the generalized answer. Perhaps someone else can chime in and build upon my answer with specific sources we have about the Goths. No doubt, they will reinforce the ideas outlined above.)