I am sorry to be pedantic about this, but I think there is a deeper point to be made here about what "completeness" can mean in terms of genome assemblies.
All the genome assemblies we have right now for various organisms are really an amalgamation of the individuals/samples/cells that were collected and sequenced. During the assembly process, the genotype that is the most frequently observed in the collected sample is used. So the resulting assembly is really a chimeric genome. Not to say this genome isn't useful, because it is incredibly useful in terms of a backbone for mapping reads to for various purposes.
So if we are referring to completeness of information about the organism, we would need to also know the population level variants seen in the genomes of individuals. There are plenty of research projects/consortiums working hard on gathering variant data (ie. 1000 genome project). There are also proposals by well regarded bioinformaticians to start utilizing graph based file formats to represent genomes so we can take advantage of the extra variant information (ie. GFA format, HISAT2 mapper).
If we are referring to completeness of the genome assembly. We are usually talking about scaffold length/conservation of genes/number of gaps type metrics that measure how well the assembler performed on the set of data given. These are more of a technical measure of algorithm performance.
You will find everything that you need in here: http://onlinelibrary.wiley.com/doi/10.1111/eva.12178/full
every draft genome assembly constitutes merely a hypothesis of the true underlying genome sequence