There is little strain on the "language leaders" (that would be barking up the wrong tree, anyway) as the features you named with one exception are already done and have been available for years.
• "is item 'in' set" operator: https://metacpan.org/pod/Syntax::Feature::In
• sets: https://metacpan.org/pod/Set::Object
• clean function parameters declaration system: https://metacpan.org/pod/Kavorka::Manual::Signatures
• constants that are not functions/objects: https://metacpan.org/pod/Const::Fast
• enums that are not functions/objects: ?
• structs that are not functions/objects: pack/unpack, for explanation see https://metacpan.org/pod/Convert::Binary::C
When I see that we still don't have a simple "is item 'in' set" operator, 12 (?) years after the introduction of smartmatch and 5 or 6 years after its half-retraction...
(Well, we still have no sets either :-), let's talk about lists/arrays...)
When I see we still don't have a clean function parameters declaration system (better than the mess with prototypes and signatures)...
When I see that constants, enums, simple structured data (C-like-ish structs) still have to be simulated with functions and/or objects and the weight which comes together...
...On one hand, it gives a set of features that could be used to justify a jump. On the other hand, if these basic features has never been deemed useful by the language leaders until now (that is, during a couple of decades), I have little hope they will change their minds.