Fix #235: bump MLKit version to 7.0.0

[mauriziopinotti]

Porting of juliansteenbakker/mobile_scanner#1206 from mobile_scanner to qr_mobile_vision.

Upgrade MLKit to version 7.0.0 to fix conflicts with new Firebase versions, see #235.

[rmtmckenzie]

I've intentionally not put in a constraint on the version of the pod as it not needed. If you're finding that this needs to be included for it to work I'd consider it, but I think that simply doing a "pod update --repo-update". My reasoning is that if someone were to be using an older version of firebase they should still be able to use the most up to date version of the plugin unless there are breaking changes that require an update to MLKit (rather than just dependency stuff).

[mauriziopinotti]

Actually, without these changes I am not able to overcome the issue described in #235

You can test it on your end by adding these dependencies in the pubspec.yaml of the example:

  firebase_core: ^3.4.0
  firebase_crashlytics: ^4.1.0
  firebase_messaging: ^15.1.0
  firebase_analytics: ^11.3.0
  firebase_performance: ^0.10.0+5
  firebase_remote_config: ^5.1.0

Then I get

[!] CocoaPods could not find compatible versions for pod "GoogleUtilities/Environment":
  In snapshot (Podfile.lock):
    GoogleUtilities/Environment (= 7.12.0, ~> 7.7)

  In Podfile:
    firebase_core (from `.symlinks/plugins/firebase_core/ios`) was resolved to 3.6.0, which depends on
      Firebase/CoreOnly (= 11.2.0) was resolved to 11.2.0, which depends on
        FirebaseCore (= 11.2.0) was resolved to 11.2.0, which depends on
          GoogleUtilities/Environment (~> 8.0)


You have either:
 * changed the constraints of dependency `GoogleUtilities/Environment` inside your development pod `firebase_core`.
   You should run `pod update GoogleUtilities/Environment` to apply changes you've made.

Running pod update GoogleUtilities/Environment asks you to run pod update nanopb that in turn asks to run pod update GoogleUtilities/Environment so I'm stuck in a loop.

Of course, I can use to my fork but it looks like other people are experiencing the same issue so I though it was a good idea to fix it upstream.