I think it would be helpful to analyze what capabilities make sense to put into the body like Star Treks Borg, (cyborgs like Robocop or Steve Austin, the 6 million dollar man), put into the cells and genetic code (like artificial X-men) and what makes sense to leave as a wearable tools (like the comic book character Iron Man).
We can also analyze what enhancements might make a difference or are flashy but mostly irrelevent. One of the main aspects of relevance would be difference in productivity and ability to earn a living of future Iron men versus future x-men versus future Borg. Note: there is no reason for the eventual appearance of any enhancement architecture to be as obvious as the fictional examples.
The analysis can also be stated as: when would it be advantageous to get closer to our technology ? When can we do just fine using technology but not becoming our technology?
One of the things that was mentioned was night vision.
Now: We have night vision goggles now which you can buy on Ebay and other places.
There is limited advantage or difference for this capability. There is less reason to out it in the body versus using the tools.
Mundane Iron Man Now: We have fork lifts and cranes.
Mundane X-men Now: We have steroids, exercise and supplements (creatine, protein powders etc...). In mice they have modified genetics for strength enhancement and I expect that it will work on humans.
More exotic: We are making exoskeletons. (more advanced Iron man and Manga style robotech/Gundam). We will perfect the genetic modifications and the techniques used to deliver those modifications.
Utility: strength helps in various situations. The main reasons for doing it are for general resilience and for an "always on" ease of use. Good exoskeletons would limit the advantage for genetically modifications in this area.
If you can make modifications that are as safe as supplements and with health benefits instead of risks, then why would you not do it? The advantage to be able to lift something is irrelevant because you could get an easily available tool or exoskeleton to help you.
Mundane now: segways, bikes, cars and planes.
Exotic now: exoskeletons.
Utility: Same as for strength. You can do it but the advantages will be limited. There would be an advantage for enhancing reflexes and reaction times. Reaction times would benefit from sensory enhancements. Being able to spot and identify targets sooner.
Mundane now: computers. PCs, iPods, there are wearable computers and displays that go directly to the eye.
Exotic now: There is Brain gate and other close interfaces between machine and brain. Many of those are non-invasive.
Utility: The difference between invasive and non-invasive is one of bandwidth and communication speed. There are also advantages to integrated control.
Productivity: This is one enhancement that could have a substantial productivity variance depending upon architecture.
This is one area where getting optimum performance irregardless of architecture will make a difference.
Life extension and regeneration:
Mundane now: drugs, vaccines, surgery, hearing aids, mechanical hearts, prosthetics, enhanced prosthetics, biosensors
Controversial now and soon: stem cells, gene therapy, iRNA
This is another where the genetic and invasive modifications are required.
It would make sense to modify our biology to make us more resistent to radiation, disease and better adapted to space and other conditions.
There are questions as to how well this would work in terms of consciousness. Eventually this architecture could diverge from the cyborg, genetic enhancement capabilities. The communication between biology and the computer and whether upgrading hybrid biology would be slower than pure computer equipment would be factors in whether architectures diverge in performance.
Related Articles and more reading:
Article in Slate talking about transhumanism
Transhuman ethics article from Reason
Transhuman reading list