To answer the questions separately:
1 - "Could the scientists tell that this man is different from other humans?"
The functional basis of DNA fingerprinting, is to use regions of DNA that are known to be highly variable between individuals, such as micro- and minisatellites (short, highly repetitive stretches of DNA).
This is explained in fairly basic terminology on the wikipedia page for the topic.
As such, if your question means "can this be used to distinguish him from other people", then the answer is yes, absolutely. That's exactly why DNA fingerprinting is used in criminal trials etc.
If the question is "could the stretches of DNA used in the fingerprinting procedure show if someone is 'super-human'" then there are 2 issues I can think of:
The first is up to you:
You have the creative freedom to decide where in his DNA these alterations have been made. If the modifications are not in the areas that DNA fingerprinting would test, then no.
Since the modifications that would give marked phenotypic changes such as sight improvement etc. are likely to need be in coding DNA sequences (the ones that typically give rise to proteins, though this is an exceedingly loose definition), it's probably pretty implausible you'd be changing regions DNA fingerprinting examines. Tandem repeats dont generally have well known or defined biological roles. They're often referred to as 'junk DNA'. Though this term is falling out of favour as we're beginning to realise there are more subtleties to genome architecture.
Secondly, if we assume you do elect to modify some satellite sites, it would be extremely unlikely that this could ever be linked to 'super human' phenotypes. Determining whether or not a mutation, even in a coding sequence, where you can examine the protein product and the organism experimentally, has any biological effect, is still extremely hard to do. Not to mention, a vanishingly small number of spontaneous mutations in a sequence ever give rise to an 'improved' sequence.
2 -"Would it be possible for someone to be human and have a different number of chromosomes?"
Absolutely, this is a well documented condition, generally called aneuploidy. 'Ploidy' is the fancy term for chromosome copy number.
Down syndrome, AKA "Trisomy 21" is probably the most well known example of this. Sufferers are very definitely still human. As the scientific name implies, they have 3 copies (so 1 additional) of chromosome 21. It is also possible to be lacking chromosomes, as well as having additional copies (for example, Turner Syndrome.
3 - "Does all DNA fingerprinting appear as bar codes and mean nothing until compared to a sample?"
Generally yes, you'd compare 2 samples to obtain an answer. You perform fingerprinting and sequence length polymorphism studies on your 'subject' sample, and on your 'reference' sample, and if they have the same banding pattern, you have a match. It will still contain information though. Those samples could be sent for sequencing for instance, which would tell you to even greater resolution, how similar 2 samples are.
Just as a final thought (though getting a bit off the scope of the SE), if you did want the authorities to be able to tell that someone had been genetically modified, you could propose that those microsatellites are changed to something non-repetitive, which would potentially stand out. Or go even further and 'watermark' the genome, much like Venter et. al did, with the first synthetic genome (they incorporated a code to spell out messages in 'junk' regions of the genome).