The seemingly highly-confused interpretation offered by the authors:
Determining polymorphisms of SNP type from chromosome Y resulted in categorizing skeleton from grave no. 1 with very high probability into haplogroup G, whereas skeleton from grave no. 2 with very high probability into one of three haplogroups J, I or E*. Detailed results of determinations are included in the attached table 2. On the basis of the above mentioned expertise one can state that the skeletons are male individuals with no relationship between each other. [. . .]
An analysis of polymorphism of single nucleotide (SNP) of chromosome Y from genetic material derived from both burials has brought in different results than in the case of so far analyzed aDNA materials of burials of the Corded Ware culture or partly contemporary Beaker culture which revealed the presence of haplogroups R1a1 and R1b among them (Haak et al., 2008; Lee et al., 2012). In case of the dead from Wrocław-Jagodno genetic diversity of both individuals was observed. One of them does not have clearly determined haplogroup. We should reject his affiliation to paragroup E* characteristic mainly for Africa and identified among population of Bantu (Karafet et al., 2008). On the other hand, haplogroup J was probably formed about 30000 years ago in Arabian Peninsula and it is often identified as a indicator of the Neolithic demic diffusion associated with spreading agriculture (Semino et al., 2004, 1996). Its contemporary distribution covers mainly the area of Middle East and the Mediterranean Sea basin; it sporadically occurs in Central Europe. Latest analyses show that its spreading might be a marker of later migrations (Giacomo et al., 2004). Hence the most probable is acceptance of haplogroup I as a proper one for the examined individual. It is considered that it was developed between 15000 and 30000 years ago (Karafet et al., 2008) and its spreading is associated with the expansion of the Paleolithic Gravettian culture (Semino, 2000) or population from the beginning of Holocene (Rootsi et al., 2004). Thus we should think that this individual is most probably descendant of native hunting and gathering community. Haplogroup G, identified in the second individual, belongs to widespread multiethnic groups of Europe, Asia and northern Africa. This haplogroup is largely identified among analyzed aDNA materials from Europe including the early Neolithic in Spain and Germany and the late Neolithic in France. It is a serious factor supporting a conception of spreading of Neolithic from the area of Middle East (Haak et al., 2010; Lacan et al., 2011; Rootsi et al., 2012). It may indicate very complicated development processes of communities of the Corded Ware culture in which diverse populations participated – autochtonous deriving from hunting and gathering ancestors as well as Neolithic populations, genetically deriving from the Middle East areas but already living there since the beginning of Neolithic.
I see no basis whatsoever in the reported results for "categorizing skeleton from grave no. 1 with very high probability into haplogroup G". Possibly there is some miscommunication among the authors, or else a complete failure of logic. It appears the authors merely failed to exclude haplogroup G in sample 1 because of an unsuccessful test (and arbitrarily decided to make haplogroup G their "very highly probable" default assumption).
At the locations where both samples returned results, the only difference is at P25 (which we'll come back to). If I and J are possibilities for the haplogroup of sample 2, then possibilities for sample 1 include I, J, and G. [Edit: I originally wrote that among these choices haplogroup I would also be my guess, and it still is; but we shouldn't have to guess, and there are still other options not ruled out beyond I, J, and G.]. Note that if we ignore P25 (as the authors appear to have chosen to do), the Y haplotypes are identical at the available resolution and provide no basis for asserting the individuals are unrelated.
I'm not sure what to make of the fact that sample 1 shows CA and sample 2 shows C at P25. P25 exists in multiple copies in a palindromic region and is not the most reliable marker. But if in sample 1 any copy of P25 is A, it would suggest sample 1 is R1b.
Against this, the authors apparently take their results to indicate that sample 1 is ancestral for M207 (which would rule out R) and M45 (which would rule out P). [Update: removed speculation about errors due to strand issues. In this case, it should be safe to rule out R as a possibility for both samples, and to rule out P for skeleton 1 (though I'd still like to see comments from the authors about the "CA" at P25 for skeleton 1). The more fundamental point remains: more data will be needed to confidently assign haplogroups to these samples.]