Quantity of languages might be less important than: how many needs are served by those languages, whether the ecosystem is dynamic enough to keep expanding served niches, and whether the culture and community is likely to produce language support for a niche that matters to you ever or on a realistic timeline. The JVM does appear to have a lot more niches covered, but you can still do all the things those languages do in what's available for the CLI.
I don't know much about the current state of CLI and .NET beyond what I've read here, but it sounds like it's dynamic enough to keep expanding. I also don't know enough about the long tail of niche languages supported by each to know which direction they're headed.
That's the situation with the tools used for music production. In theory, any DAW (Digital Audio Workstation) can make any kind of music. In practice, they all move toward different kinds of music, and you'll run into increasing friction as you do weirder or more complex stuff if you pick the wrong DAW. Cubase can do electronic music, but you're better off with FL Studio or Live. Live and FL Studio can do orchestral, but you're better off with Cubase.
And I'd guess there's a similar dynamic with CLI and JVM and the languages that target them.
I am told that the C# language has evolved faster and better than Java. Despite its origins in .NET 1.0 being basically a "Microsoft Java".
But paradoxically there is now more diversity of actively used languages that target the JVM. e.g. Kotlin, Scala, Clojure.