Dr. Thomas Lee of Stanford University, has written a very nice article outlining the development of the Analog design. In this article, reference is made to the use of positive feedback by Bob Widlar in the design of the world's first commercial Solid State OPAMP, the uA702. A similar kind of positive voltage feedback is used in all the Philbrick Tube OPAMP modules.
See K2-W, K2-X, K2-XA, K2-YJ, SK2-V for the positive voltage feedback resistor between the cathode follower output and the middle stage cathode. In addition, the SK2-V also has cross coupled positive feedback around the second differential pair of triodes (the 12AX7's).
The primary reason to use positive feedback in these tube OPAMPs and in the Solid State uA702 is that high impedance current source loads are not available, or would be prohibitive to implement with tubes. Positive voltage feedback is a way to subtract the unwanted conductance of the resistor loads. This method has the welcome advantage of not adding poles to the loop response, as would be the case with additional gain stages. With positive voltage feedback, DC Gain is increased with a negligible effect on unit gain bandwidth or stability.
The article was published in the October 2007 issue of The Solid State Circuits Society News. You can reach the SSCS main page and look for the article
"Tales of the Continuum: A Subsampled History of Analog Circuits
Thomas H. Lee, Center for Integrated Systems, Stanford University"
In October 2007, the direct link to the article is here, but it may change as the October 2007 issue is replaced.