@pieter While the discussion about data persistence is interesting, I’m not sure this addresses what I take to be your actual goal, namely getting a list of plants in Tanzania. The data set that has vanished only included material in the National Herbarium of Tanzania, but I suspect that data on Tanzania is distributed across many herbaria and hence any one institution is unlikely to have a full list of species.
Have you tried New feature: download lists of distinct species contained in occurrence searches? This should retrieve a list of all taxa recorded in Tanzania (you can restrict the search to whatever taxonomic group matches your definition of “plants”). One drawback is this list will include non-native species as well. Perhaps you could compare the lists here: Search. Another way to check might be via the Plants of the World https://powo.science.kew.org which has data on whether species are native to particular areas.
Apologies if you’ve thought of all this already. I think it’s one of the biggest weakness of GBIF is that it often fails to make it easy to answer fairly straightforward questions (or at least, questions that feel like they should be straightforward).