> why dead fragments can't join again to make full blown virus?
Is a good question.
Around 220 species of viruses are known to be able to infect humans. Another 8 at least can affect dogs, but not humans. Lets say that 50 more viruses are exclusive from each one of dogs, cats, cattle, pigs, poultry and fishes and most main crops
Those species are studied. Less imagine that this list makes a total of 5000 viruses relatively well studied.
Well. In a conservative estimate, only Mammals could host more than 300,000 species of virus still unknown to science. There are, probably, several millions of virus still unknown only among arthropods.
Lets say that "1%" of the viruses don't work that way, and we don't know anything about the regenerating DNA capacities of the remaining "99%", but we can assume that probably will not happen. Probably. Viruses stop being functional when devoid of its capsides.
But... I wouldn't bet against viroids, in this sense. Fortunately, we aren't plants.