I think this is an important fact to have in mind for authors who are trying to decide whether or not to put their books in the KDP select program. If the vast majority of US ebook customers are on kindle unlimited, then it makes it more advantageous for authors to put their books there.
Amazon doesn't reveal stats like this.
However, in 2017, The Written Word Media did an estimate based on the KU pot payout to make an estimate, which came out at 2.5-3 million readers:
2.5 million readers in KU is a lot of readers, and because our calculations are based on pages read it’s likely that the 2.5 Million number represents the active readers enrolled in KU. We would guess there are even more inactive users who are subscribed but are not reading. There are limitations with our data, and we’re making quite a few educated assumptions but we think it’s safe to assume that there at least 3 million readers in Amazon’s Kindle Unlimited.
Not an answer to your question but too long for a comment. I beg for tolerance from the moderators ...
How do you see the number of KU users affecting your decisions? There's nothing exclusive about KU. I have several books being sold as paperbacks, as Kindle "orindary sales", and also in Kindle Unlimited. If the number of users is very small, then whether you put you book in KU or not is close to irrelevant. If it's large, then whether to enroll is an important question, but I don't see how the number of users changes the calculation.
Each person who reads your book on KU is a lost potential sale. If your book wasn't available on KU, they might have bought a copy. On the other hand, if it wasn't available on KU, they might have skipped it completely and read something else. So yeah, you get less per reader on KU, but in the long run are you gaining or losing? Note that the number of KU users of course changes the magnitude of the answer, but not the direction. If, say, each KU reader gives you $2 and each straight purchase gives you $4, then if every KU reader costs you one sale you are losing $2 per KU reader. If half the KU readers would not have bought your book, then it's x * $2 versus x/2 * $4, i.e. you break even. If less than half of KU readers would have bought your book, then you are ahead by using KU.
This is further complicated by considerations like, maybe being on KU will get you a lot of exposure that you wouldn't have gotten otherwise, and the KU readers will tell others about your book, and some of those will buy copies. Or they'll write reviews that will boost sales.
I don't know if anyone has collected and analyzed statistics to give any sort of useful answer.
As an author whose books sell hundreds of copies -- not millions, not thousands, hundreds -- I figure that any exposure is good, so I sign my books up for KU.