WontFix as this it is working as intended.

wrapKey() with 'jwk' is expected to return an ArrayBuffer, NOT a JWT:
https://w3c.github.io/webcrypto/Overview.html#dfn-SubtleCrypto-method-wrapKey

'jwk' in this case simply tells it what format to export the key in prior to encryption.

If you are unhappy with this behavior best to take it up with the spec:
https://github.com/w3c/webcrypto/issues

Hmmm , but the spec you link to says:
====
If format is equal to the string "jwk":
Convert key to an ECMAScript Object, as specified in [ WebIDL].

Let json be the result of representing key as a UTF-16 string conforming to the JSON grammar; for example, by executing the JSON.stringify algorithm specified in ECMA262.

Let bytes be the byte sequence the results from converting json, a JavaScript String comprised of UTF-16 code points, to UTF-8 code points.
==
Also in the spec https://w3c.github.io/webcrypto/Overview.html#subtlecrypto-interface-datatypes
"jwk": The key is a JsonWebKey dictionary encoded as a JavaScript object
====

That is step 14. And then in the following step (15) you call the wrapKey() or encrypt() operation on that, which outputs BYTES.

So let's say you are doing:
  wrapKey('jwk', key, wrappingKey, wrapAlgorithm='aes-cbc')

Conceptually this boils down to:
  AesCbcEncryptBytes(wrappingKey, ExportKeyToJsonBytes(key))


The step 14 is that ExportKeyToJsonBytes(key) expression above.


The output from wrapKey() is always going to be ciphered bytes.
Certainly it wouldn't make sense for the output to be a JWK.

As far as a JWT, isn't that just a delivery vehicle for signed data? That wouldn't be appropriate here either, but I suppose it could be crammed into a JWT.

OK I guess that's an impedance mismatch between JWK RFC and the W3C spec
For instance https://tools.ietf.org/html/rfc7520#section-5.8

Ultimately what I want to make sure is that it can be make cross platform / cross technology.

Ah, so your expectation was to get a JWE out?

JOSE is not my area of expertise; I suggest asking questions either on the webcrypto spec github issues (https://github.com/w3c/webcrypto/issues), or on the WG's mailing list (public-webcrypto@w3.org).

That said there were numerous discussions about the interop of these standards, so I don't expect this is going to be changed.

There are no plans for the Chrome team to support JWE, and that was intentional and, as  Eric mentioned, discussed at great length. The public-webcrypto archives have extensive documentation of the 2+ years of discussion about JOSE specs, for which the final decision was that WebCrypto is *NOT* a JOSE API and it does not emit JOSE objects.