Comparing various means for rehabilitation one will invariably have to weigh one issue against the other. And mind you - a lot of the problems are not apparent at the desk, or in theory. A lot of the nasty and sometimes long term problems manifest themselves only in blood, flesh, juice and when putting things to action.
The following overview and table (below) give an overview over my current rating of various aspects. These relate to options regarding what to do with the damaged arm.
Looks. Obviously, it matters how it looks. However, many people don't know that no matter what I wear on my stump, people will find something to object, and no matter what I wear, they will reject me, or find it all relatively stressful. These are normal reactions as we know. But they limit what I can achieve even with advanced means of silicone cosmesis. In other words, it only matters to others to a very small degree what type of appearance my arm has - as long as it is clean and well groomed, washed and soaped down.
Rejection rate. Interestingly, despite a rather anatomic appearance, many prostheses are rejected by amputees. The bare stump or a Krukenberg stump (even more) rate surprisingly high with amputees which, when examining acceptance / rejection rates per se (which is hardly ever done), may come as a surprise. True orthopedic specialists also confirm that the Krukenberg arm is functionally superior to any other option  as opposed to people that sell prostheses and thus are neither interested in stump or Krukenberg based functionality .
First of all, bare skin is deformable and has superb grip characteristics. Any similar setup on a prosthetic hand will have you re-do the grip surface cover every 3-6 days. If you can get the materials and find time for that, great. If you can have self-healing skin instead, greater.
Secondly, even with a stump without particular grip function, "manualization" of the body and surrounding still allow for a relatively high degree of function.
Over decades, asymmetry becomes a determining issue that can have serious consequences. So to keep in mind that 20 or 30 years of not wearing a prosthesis and just using the stump can be problematic, one may want to consider alternatives. But in terms of acceptance, obviously what is one's own is better - be it a stump or one that has been modified to increase grip function (Krukenberg).
Cost is not just financial. Cost is also emotional. Every time my prosthetic arm fails and the broken or torn off parts injure me, I suffer. Each time I need to take out time to go there to get it repaired. The maintenance required. The upkeep. Insurance letters. Cost of fitting, maintaining and running a prosthetic arm is considerable. Compared to that, natural alternatives just shine.
Complications that can occur are relevant to consider. Particularly, long term complications are a real issue.
Most people forget that wearing and actually using and heavily loading prosthetic arms with any given or absent degree of functionality will entail complications as necessary consequence. Logically, a number of current review articles  are obsolete as they completely ignore this aspect and from that, we can conclude that even these authors are absolutely clueless inasmuch as that subject is concerned, or, they make their living by selling prosthetic arms which then will constitute a clear bias and conflict of interest .
A non functional or amputated arm obviously results in overuse and asymmetry. But a heavy prosthesis can contribute to serious shoulder problems, and due to its weight might be also under-used and thus contribute also to overuse of the remaining arm and hand. In other words one will want to find a good functional and weight wise balance - regardless of how it looks like. In terms of real activity, users of body powered arms lead the statistics to no surprise - even my little point score overview shows this to be a rationally founded result.
Skin issues are mostly not too much of a long term problem if one manages a good hygiene and does not overdo it. My skin problems stopped mostly after about 1-2 years, as I now found ways to keep things under control. In reality, doctors will lose most amputees out of sight before long term complications start. Often, prosthetic users lie to doctors about how well they use the prosthetic arm as many doctors cannot take no for an answer and are thus not trustworthy enough to be entrusted with the bad news, which is, that the clunky device is not nearly as useful as everyone had hoped it to be.
Grip function obviously is pivotal. If that is not good, the whole body "manualizes" around the objects and while it can be good exercise for a day or two, it will be bad over a decade or two.
It is interesting to critically play with these factors and aspects. Using a conservative and cautious point score the result here is a surprise, it really is. I wonder whether research is being done that improves choices and weights used for this table.
|Technical measure||How other people think it looks||Rejection rate with amputees||Cost||Complications||Grip functionality||Sum [points]|
|Stump||Some like it, some don't [3/5]||50% like it [2.5/5]||No extra costs [5/5]||Asymmetry, dysbalance, overuse of other arm [1/5]||Not good [2/5]||11.5|
|Krukenberg stump||Most don't like it on photos, but 50% or more like it when they see it in action [2.5/5]||Very high acceptance rate [4/5]||Low cost (operation, physiotherapy) [4/5]||Some asymmetry but best functional result of all that are available [4/5]||Best of all [5/5]||19.5|
|Cosmetic prosthesis||Usually well accepted [4/5]||30% acceptance rate [1.5/5]||Low cost (prosthesis) [3/5]||Asymmetry, dysbalance, overuse of other arm [2/5]||Bad [1/5]||11.5|
|Body powered prosthesis||Usually well accepted [3/5]||30% acceptance rate [1.5/5]||Relatively moderate cost (prosthesis) [2/5]||Moderate asymmetry and overuse [3/5]||With exchangeable terminal devices and functional parts, not too bad [4/5]||13.5|
|Myoelectric prosthesis||Usually well accepted [3/5]||30% acceptance rate [1.5/5]||Very high cost and maintenance efforts (prosthesis) [1/5]||High asymmetry, prosthetic weight, moderately to extremely uncomfortable socket and overuse [2/5]||Not too bad but not great either [2.5/5]||9.5|
As the subversive mind that I have I see another use for that table. It is undisputably (I have not even started to lay out all arguments and references yet) an extremely well researched evaluation that I came up with here.
However, you will find any orthopedic surgeon to tell you that myoelectric prostheses are best and Krukenberg arms are obsolete. Given a far different reality (see above), here is my question: what aspects do you necessarily need to ignore in order to arrive at such a conclusion? What mindset do you need to have in order to weigh myoelectric arms superior?
The answer is clear and I refer to the columns of the table above: you need to weigh "how other people think how it looks" very highly; you need to necessarily ignore acceptance rates of other amputees; you definitely will ignore the patient's own time and finances by recommending a prosthetic solution; you will not consider long term problems that are to be expected with prosthetic arms, and grip functionality has a limited significance for you. However and first and foremost, you will weigh some "luxury item / gadget" aspect that I did not even list here because it is laughable. If you put emphasis on looks and extremely weigh gadgetry, and if at the same time you ignore the other factors (that are undeniably there) - only then a myo arm will win.
And that is the strength of a good analytical layout. It can show you how to look at things differently.
 Sachin Watve, Greg Dodd, Ruth MacDonald, Elizabeth R. Stoppard, Upper limb prosthetic rehabilitation, Orthopaedics and Trauma, Volume 25, Issue 2, April 2011, Pages 135-142, ISSN 1877-1327, DOI: 10.1016/j.mporth.2010.10.003.
 Baas, N., Schmidt, A., Bühren, V., Öhlbauer, M., Giessler, G. (2009) Traumatische Amputationen an Unterarm und Hand - Trauma und Berufskrankheit. 11(1): 7-16. http://dx.doi.org/10.1007/s10039-009-1476-3
Doi: 10.1007/s10039-009-1476-3 http://www.springerlink.com/content/23m655067571577v