I have not had any direct experience with pure cobalt. I do have some limited experience with cobalt alloys, such as Stellite (chrome and carbon in cobalt). I would assume that the reactions you are getting from others (too brittle, etc.) are based on their knowledge of cobalt alloys rather than pure cobalt. Here are some facts about pure cobalt (99.9%) [taken from ASM Metals Handbook, 9th ed., Vol. 2, p 725-726], which are the basis of some of my tentative opinions on this matter:
Tensile strength in the as-cast condition: about 34 ksi (thousand pounds per inch) [note that this is about 1/2 to 2/3 the value for mild steel, which would be about 50-60 ksi, depending on the grade)
Tensile strength in the annealed condition: about 37 ksi (i.e. not much change)
Tensile strength in the swaged condition: about 100 ksi.
Pure cobalt has an HCP crystal structure.
Melting temperature of 2723 deg F (1495 deg C).
Electrical conductivity is about 28% of pure copper
Thermal conductivity is 69 W/m2K, i.e. about 20% of pure copper.
Specific heat is 414 J/kgK, i.e. about the same as copper and steel.
From the above, I draw the following conclusions:
1) The material can be cold-worked to a high degree, based on the information about tensile strength in the swaged condition.
2) The material is likely to be ductile, since it is an essential pure HCP crystal structure plus the large amount of swaging that appears to be possible. Brittle alloys would not show such increases in strength upon swaging (you really cannot swage them at all...think cast iron as an example).
3) The melting temperature, electrical conductivity, specific heat, and thermal conductivity show similarities to steel alloys, i.e. there is nothing highly unusual that would prevent using conventional welding techniques.
Of course, the question is what would you weld it with? Could you do this weld autogenously (i.e. without filler or using excess material as the filler)? I do not know how the material would react but it seems to be a reasonable possibility. Looking at a couple of phase diagrams, it appears that either pure copper or pure nickel could be used with a resulting weld metal that might be acceptable, but this is just a best guess. Any way you do this, the weld is likely to be somewhat harder and less ductile than the pure cobalt. Also note that these two fillers are pure copper and pure nickel (an "unalloyed" nickel filler metal is commercially available, see Inco's Nickel 200 filler metal, but it may be hard to get routinely). Use of a conventional filler metal (such as the Monel filler mentioned by one of your sources) could lead to some unpredictable results because of the interactions between all the alloying elements. As a guess, I think you could probably use a copper braze material also.
For any of these cases, welding up a test sample would be highly recommended.
Tom Doody (remove NS from email address above for direct contact with me)