It can happen. We expect a duplicated gene to either evolve to a slightly different function or degrade to a pseudogene, but there may be circumstances where this does not happen. Maybe the duplication was so recent in evolutionary history that it just hasn't happened yet. Maybe there's an evolutionary advantage in keeping multiple copies - maybe to allow for massively increased transcription. Your example is a histone protein which is needed in great abundance in the cell.

There is nothing wrong or unusual here. We are looking at a gene family (histone cluster) whose members will be pretty much related (or even identical). So we have gotHIST1H4F, HIST1H4I, etc, all similar members of the same family. If you look at HIST1H4Fin Ensembl you can see that it's been cross referenced against all the others members of that family

http://www.ensembl.org/Homo_sapiens/Gene/Matches?db=core;g=ENSG00000274618;r=6:26240426-26240737;t=ENST00000244537

The mappings are ok. This is in inherent feature of this given gene family.For the two examples above the protein sequences are 100% identical:

http://www.ensembl.org/Homo_sapiens/Transcript/Sequence_Protein?db=core;g=ENSG00000274618;r=6:26240426-26240737;t=ENST00000244537

http://www.ensembl.org/Homo_sapiens/Transcript/Sequence_Protein?db=core;g=ENSG00000276180;r=6:27138588-27139881;t=ENST00000615353

The are all paralogs. They are different loci (at different positions in the genome) but code for the same protein. Look for them in the Ensembl gene tree below (highlighted in blue):

http://goo.gl/SMBbIF

Yep. It's certainly possible. The human genes UBB and UBC each encode a few tandem copies of the same ubiquitin peptide sequence (unless I'm mistaken); hence, any given cellular ubiquitin peptide could have come from either of the genes (or from one of the other ubiquitin genes)