and did you check cbitcoinaddress, cbitcoinaddressvisitor, cbase58data, etc. that there is nothing hidden that would let the cbitcoinaddress constructors or "lostWallet.Get() == addressSource.Get()" do something unexpected? you can confirm that the == operator works correctly and has no tiny flaw added that will fail in this special case?
it's easy to quote a few lines of code that somehow look like that everything is ok while there are several thousand lines of code involved in making this one comparison work and where you could easily hide your stuff.
Coding pros make use of tools like windiff and winmerge. That is the first thing I did even before applying the patches..
I also did negative whitebox testing, using owned addresses..
Is it enough? maybe not..
But at least read the actual thread to understand what has been done.. and/or look at the commit history.. and/or do your own diff/audit.. before making these uneducated comments.