Hmmm. That's a tough question. You only have to go back 30 years and you will already find a lot of outdated methods/concepts or even findings. There are a lot of older works which still contain a lot of relevant information and are often still refered to. If you are talking about displaying the full range of historical methods and concepts we currently consider near indispensible, I don't think you can go back much further than 1970 to 1980 without noticing glaring differences in approach. This mostly goes for qualitative research.

However, if you are looking for solid historical quantitative research, there have been some historians who did some solid groundwork shortly after Leopold von Ranke. So that's the early 20th Century basically. I'd be extremely wary of referencing qualitative information from this period, but there is a lot of good quantitative information to be found there. I've read some treatises from the 19th Century which were surprisingly tought-through and that were put together very well, but they obviously lack a lot of concepts we consider crucial. If you want an example of some of those concepts, just look at the subreddit rules, many of the rules in there are considered essential in historiography and were absent in these works.

So I guess it really depends what information you are after, but I'd be very critical of anything prior to 1970.