Investigate the implications of 964070

[arlolra]

@aleth suggests we have a look at 964070 and see how it interacts with OTR, and whether to recommend a different default value for the pref.

/cc @azadi

[arlolra]

The patch just adds another call to writeMessage(), which is the correct method for the transformation pipeline, so no problems there.

As xnyhps says, instance tag support should help with OTR messages ending up at the wrong device. See #45 for that.

However, as the XEP suggests, it might be nice to add an API in IB to append <private/> and <no-copy/> elements to <message/>s in an OTR session.

It's probably ok to leave this enabled by default as it stands, though there's the potential for some poor interactions.