That is a serious "quotation needed" statement. How can you even know you know about all relevant ecosystems when you are missing 90% of species? Besides, just knowing about the "keystones" of an ecosystem is required but not sufficient.
As an analogy, if you knew the top 10% keystone chemicals that make up the human body but did not know the long tail of the other 90%, you'd likely miss out on a lot of the details that make it all work like immune system cells and vitamins. Just because a molecule (or species) is rare does not make it non-essential.
I hope we are building a statistical model of which one is eating which to map where are the missing pieces of the puzzle and direct our focus of research.
From The Diversity Of Life by Edward O. Wilson
(These numbers are for recognized species.) This does not answer the "only 10% known" part. However, the part where the author says 13000 new species recognized every year should give an idea. Also this is about species not members under those species. And from the same book "Bacteria continue to be the “black hole” of biodiversity, their depths unplumbed."
"How many species? The estimates I made in 1986 and cited in The Diversity of Life (1992) put the number of recognized living species in the world— in other words, those formally described and bearing two-part scientific names— at approximately 1.4 million.
About 13,000 additional “new” species are recognized each year. Thus in the decade since 1986 more than 100,000 species have been added, bringing the total as I counted it to 1.5 million. Meanwhile, the numbers in some important groups have been revised upward— in particular in the insects, the largest of all groups, from 751,000 in the 1980s to 865,000 in 1998. A similar elevation has been made for the fungi, from 47,000 to 69,000 species. A commonly cited total world figure in the late 1990s, suggested for example in the Global Biodiversity Assessment of the United Nations Environment Programme (1995), is 1.75 million species. But this does not take into account the number of formal species names that have been erroneously applied to species named by earlier investigators, requiring an eventual reduction of the global number by 10 or even 20 percent...
“Working figures” for all groups, including insects, have tended to fall close to 10 million; the Global Biodiversity Assessment number, which constitutes no more than an educated guess and leans to conservative sentiment, is 13,620,000."