Consumer: Fix sequence number gap

[jmillan]

Drop the given sequence number if the RTP packet cannot be forwarded.

Fixes #1069

By not doing so, we are incorrectly generating non legitimate sequence number jumps which involves:

• The other end requesting NACKs for packets that do not exist, and us processing them.
• The other end waiting for those packets that will never arrive.
• Possible implications an jitter buffer and decoding level.

This problem is very apparent when using BWE and the available BW is low enough to discard a lot of packets, or when the outgoing bitrate is manually set low enough too.

[ibc]

I'm amazed that this fix looks too obvious to never have realized of it. I mean, we were literally dropping packets in MANY SCENARIOS without dropping then in SeqManager. How is possible that we never noticed it?

Please add entry in CHANGELOG.

Also minimal cosmetic changes.

And ASAN checks are failing in CI.

[ibc]

Fixes #1069

Fixes #1069 or fixes #1070 or both?

[ibc]

I'm updating the branch with pending changes in v3.

[ibc]

And ASAN checks are failing in CI.

All those are caused by the new TestSimpleConsumer.cpp file.