Quaternionic Kähler manifolds with Hermitian and Norden metrics (Q1945669)

This is an interesting but somehow very confusing paper. The author is interested in: {\parindent=0.5cm\begin{itemize}\item[1.] almost hypercomplex structures with pseudo-Riemannian structure \(g\) such that \(g(J_i\;, J_i\;)=e_i g(\;,\;)\) with \(i=1,2,3\) and \(e_1 =1 , e_2 =e_3 =-1\). \item[2.] in quaternionic Kähler manifolds \((M,Q,g)\), i.e., \[ (\nabla _x J_i)y=\omega _k (x) J_j y-\omega _j (x) J_k y , \] where \(Q\) is generated by three almost complex structures \(J_i\) and \(\omega _i\) are three locally defined 1-forms, \((i, j, k)\) as a permutation takes the same order as \((1, 2, 3)\). \end{itemize}} The formula in 2. was copied from [\textit{D. V. Alekseevsky} and \textit{S. Marchiafava}, Ann. Mat. Pura Appl., IV. Ser. 171, 205--273 (1996; Zbl 0968.53033), page 244] (see the beginning of Section 2, page 522), where it was called a quaternionic Hermitian structure instead. Here, by an almost hypercomplex structure is meant a set of three almost complex structures \(H=\{ J_1 , J_2 , J_3 \}\) such that \(J_i J_j =-J_j J_i =J_k\) with \((i, j, k)\) having the same order as \((1, 2, 3)\). The author calls \((g, J_1 )\) an almost Hermitian structure, but it actually is only a pseudo-Hermitian structure with signature \((2n, 2n)\). It was called a pseudo-Hermitian structure in [\textit{K. Gribachev} and the author, J. Geom. 88, No. 1--2, 41--52 (2008; Zbl 1138.53040); \textit{K. Gribachev} et al., in: Proceedings of the 6th international workshop on complex structures and vector fields, St. Konstantin, Bulgaria, 2002. River Edge, NJ: World Scientific. 51--62 (2003; Zbl 1072.53526); the author, in: Topics in almost Hermitian geometry and related fields. Proceedings in honor of Prof. K. Sekigawa's 60th birthday, Niigata, Japan, 2004. Hackensack, NJ: World Scientific. 170--185 (2005; Zbl 1109.53050); the author and \textit{K. Sekigawa}, in: Contemporary aspects of complex analysis, differential geometry and mathematical physics. Proceedings of the 7th international workshop on complex structures and vector fields, Plovdiv, Bulgaria, 2004. Hackensack, NJ: World Scientific. 174--186 (2005; Zbl 1218.53034)]. The structures \((g, J_j )\), \(j=2, 3\), are called anti-Hermitian metrics in [the author, J. Geom. Phys. 61, No. 1, 248--259 (2011; Zbl 1229.53056); the author and \textit{K. Gribachev}, Int. J. Geom. Methods Mod. Phys. 8, No. 1, 115--131 (2011; Zbl 1214.53056)]. They are more like the real part of the holomorphic symmetric 2 differential tensors. Anyway, the Hermitian structure is quite different from that in [\textit{A. Gray} and \textit{L. M. Hervella} Ann. Mat. Pura Appl., IV. Ser. 123, 35--58 (1980; Zbl 0444.53032)] but more like the one in [\textit{G. Ganchev} and \textit{A. Borisov}, C. R. Acad. Bulg. Sci. 39, No. 5, 31--34 (1986 Zbl 0608.53031)]. Therefore, [Zbl 0444.53032] does not apply to this case since it used \(\mathrm{U}(4n)\) representations. So is [Zbl 0968.53033], since it used a \(\mathrm{Sp}_1 \times \mathrm{Sp}(n)\) structure, which is compact. This paper is a continuation of [Zbl 1138.53040] to [Zbl 1218.53034], which dealt with the case of 1. The author puts his effort on the additional condition 2. He is trying to prove that these kind of manifolds are Einstein if the dimension \(4n\geq 8\) (Theorem 2.7). He also tries to pinpoint the non-general-non-hyper-Kähler case regarding the first derivatives of the almost complex structures. That is where the author uses the fundamental results from [\textit{G. Ganchev} and \textit{A. Borisov}, C. R. Acad. Bulg. Sci. 39, No. 5, 31--34 (1986; Zbl 0608.53031)] and [Zbl 0444.53032]; see Proposition 3.5 and Conclusion 3.7. In that case, the author claims that \((g, J_1 )\) is Kähler (Proposition 3.5), \(\omega _2 = \omega _3 = 0\) (proof of Proposition 3.5), being Ricci symmetric (\(\nabla\mathrm{Ric }= 0\) in Proposition 4.3, actually Einstein!), being flat iff the local 1-form \(\omega _1\) is closed (Proposition 4.2). Lemma 2.2 basically means \[ d\omega _2 = \omega _1 \wedge \omega _3 , \] and \[ d\omega _3 = \omega _2 \wedge \omega _1 \] and similarly for (13) and (18) as well as Proposition 2.9. The reviewer hopes that these flaws could be recovered in the future.