A list of MANDATORY recordings that doesn't include a single one of Bruno Walter's contributions? But does include Klemperer? Walter knew and worked with Mahler a lot longer than Klemperer ever did and we can hope that Walter absorbed some of Mahler's ideas and personality. We'll never know. Every time I listen to the fifth in old, crackly mono from NY I am pleasantly surprised just how thrilling, driven and "modern" that recording sounds. It's really rather interesting comparing Walter & Klemperer in recordings of the 2nd, 9th, and DLVDE - so different yet both so vital.
I think there has been a misunderstanding. As I specified before jotting it down, I considered complete and almost comple cycles of symphonies. I added Das Klagende Lied, Das Lied von der Erde, the Cooke's M10 and Lieder separately because they are too often (and I thik wrongly) not included in the boxes of the cycles (despite being so varied, Mahler's opera omnia is very consistent, compact and organic) and they should really be known alongiside the symphonies.
In a list of single, important releases, Walter would pop out for sure, as would Karajan and many others.
As for the fact that Walter's Mahler must be more reliable than Klemperer's because of the many years spent by him as Mahler's assistent, friend and confident, we can't be entirely sure. Walter conducted only one of Mahler Symphonies (the Third), during Mahler's lifetime and it did it after Mahler had left Vienna. So, we do not know what Mahler thought of Walter's Mahler. According to contemporary witnesses that had attended Mahler's and Walter's concert, they differed in style and approach, Walter being milder, less provocative, less extreme. On the other hand, Berlin critics during the 20s, considered Klemperer the true heir of Mahler, he reminded them of the same approach to conducting of Mahler. Moreover, especially during the 30s, the conservative Walter had tried to impose Mahler's music as the official music of the new, postwar Austria, that is to say the music in which the new nation could identify itself, the music in which it was mirrored the nation's image of the glorious past and its authoritathive legacy brought in the present. The present of the Austria in the 30s was the catholic, right-wing government of Schussnig, a friend of Walter, who was present in his official capacity at the concert for the 25th anniversary of Mahler death. So, Walter tried to make sound Mahler's music less disruptive, more conservative.
Let's not forget he was a prominent public figure: the Nazi wanted to put him away and he saved his neck because he was in Amsterdam during the Anschluss.
Consider also his attitude towards Mahler's symphonies: he never conducted the Sixth, the Seventh he conducted once. He performed regularly Das Lied and the Ninth above all for a sense of duty, because they had been entrusted to him, Das Lied in particular. After he fled Austria, apart from the first recording of the Fifth (which I like), he conducted above all M1, M2 & many M4s and refused the offer to record an entire cycle. Conductors, let's say, more "progressive" (like Mitropoulos) had no problems in conducting M3, 5, 6, 7, that is to say the more modernistic symphonies.
Another example of unreliability of assumptions about master-disciple relationship is represented by Oscar Fried. He was the first to record the Second Symphony at the beginning of the 20s. The first time he conducted it, he went to Vienna to study it at the piano with Mahler. Mahler was pleased. In Berlin at the dress rehearsal, Mahler was however very surprised by the fact that Fried seemed to have forgotten all that was said. In fact, he did the opposite! Mahler protested and Fried had to sort of improvise a new approach during the actual concert (the same concert in which Klemperer was conducting the offstage orchestra and made his first encounter with Mahler). Now, how can we be sure that the M2 he recorded in the 20s is a M2 that Mahler would approve?
Let's think about the example of Klemperer once more. As mentioned above, during Klemperer's most successful years in Germany, people who attended concerts of both Mahler and Klemperer considered Klemperer the true heir of Mahler. We also know that, in Prague, Klemperer attended all the rehearsal of Mahler's Seventh Symphony. He witnessed Mahler's work with the orchestra, his labours in revising and polishing the score, he even sat at the same table for dinner every night, listening to all he had to say about music and many other things. It's then fair to assume that he knew very well the Seventh and what Mahler wanted to achieve in performing that symphony. Sixty years after, Klemperer recorded the Seventh. It is a performance more than a bit strange, nevertheless once in while I like to listen to it because it fascinates me, but, that said, just considering Klemperer's tempos, we can be sure this is not how Mahler conducted it.
It is simply not in the nature of great artists with strong personalities like Walter and Klemperer just to reproduce the work of their mentor, they must have it done their own way and we have to judge their work above all for what it is, for its own merits, for what they want and have to say for themselves, not just for the kind of human relationship they enjoyed with Mahler.
(Sorry for the length of the post.)