Interpolation of bilinear operators and compactness (Q975851)

The authors analyse the behavior of bilinear operators acting on interpolation spaces generated by the general real method in relation to compactness. They prove analogs of classical compactness theorems (in the linear case) due to \textit{J.\,L.\thinspace Lions} and \textit{J.\,Peetre} [Publ.\ Math., Inst.\ Hautes Étud.\ Sci.\ 19, 5--68 (1964; Zbl 0148.11403)], \textit{K.\,Hayakawa} [J.~Math.\ Soc.\ Japan\ 21, 189--199 (1969; Zbl 0181.13703)], and \textit{A.\,Persson} [Ark.\ Mat.\ 5, 215--219 (1964; Zbl 0128.35204)]. However, the authors neither mention that, due to \textit{M.\,Cwikel} [Duke Math.\ J.\ 65, 333--343 (1992; Zbl 0787.46062)], the linear case was solved in full generality, nor explain why they cannot provide such a general result for the bilinear case; recall that Cwikel deduced the full general result from Hayakawa's result (where not only one, but both restrictions are assumed to be compact) by reiteration and dealing with \(K\)-functionals -- it would have been nice to see why these techniques fail (?) when passing over to the bilinear case.