首先这里是such A + as引导定从。倒不是完全割裂的。
这点你的判断很对，as在这确实引导定语从句，可以参看一下语法书中的相关内容。
柯林斯字典中有这句话：
Britain is not enjoying such prosperity as it was in the mid-1980s.
字面的理解是：英国已不再有（它在20世纪28年代中期的）繁荣。括号里为定语从句。
长难句分析中若图方便，确实可以直接看成such A as B（如B一样的A）的结构，上述这句话可以理解为：英国已没有如20世纪80年代中期那样的繁荣。个人猜测，杨鹏老师书里可能出于这个目的把它归为这个结构，确实不够严谨。

其次，你的译文大致没什么问题，就定从的翻译有一个小建议：定从本质是修饰成分，不要把它变为主干，有些时候可能出现理解侧重点偏差。建议翻译为：已经展示在神经细胞里的XXX

这里再补充一下对such A as B这个结构我个人的看法
as B部分无论是as sth.还是as引导定从，本质上都是对A的修饰
所以把such A和as B割裂开来看不是不可以。
为了便于理解，语法书也好，长难句分析法则也好，都会倾向把它们看作一个整体。

其实我主要是针对杨鹏的such thing as something结构来提出不同意见的，as后面不是跟something，是跟定语从句，而且as在其中是主语。

such ... as ... 结构中的 as 也可以引导从句，不是只能作介词接名词性短语。杨的分析没有问题。

你那段中译，
已表明楼主对句子理解出了偏差。
你说such与as句无关也缺说服力

BTW,
技术文章，这句子非常一般，我根本看不出“长难”。而且，这句子写得一般，不算很好。
对于不熟技术语体、不懂correlation方法的师爷，讲解这类语言，真是难为他啦

按近几年的写法，加两个dash 便于阅读，没问题。其次，correlation想GG明确，加them进去。
.
Such variations － in size, shape, chemistry, conduction speed, excitation threshold, and the like as had been demonstrated in nerve cells－remained negligible in significance for any possible correlation 【 between them and 】 the manifold dimensions of mental experience.

Such variations in size, shape, chemistry, conduction speed, excitation threshold, and the like as had been demonstrated in nerve cells remained negligible in significance for any possible correlation with the manifold dimensions of mental experience.
我不会用那么多术语去分析这句句子，但要理解却并不困难：
[Variations in size, shape, chemistry, conduction speed, excitation threshold, and the like had been demonstrated in nerve cells.] [Such variations remained negligible in significance for any possible correlation with the manifold dimensions of mental experience.]
我理解的大意是：神经细胞里显出了各种变化，然而一个人的内心感受繁多，即使这些变化与其有任何关连，仍然是微不足道。
【仅供参考】

阅读理解的原文如下（biying搜索到的）：
By 1950, the results of attempts to relate brain processes to mental experience appeared rather discouraging. Such variations in size, shape, chemistry, conduction speed, excitation threshold, and the like as had been demonstrated in nerve cells remained negligible in significance for any possible correlation with the manifold dimensions of mental experience.
Near the turn of the century, it had been suggested by Hering that different modes of sensation, such as pain, taste, and color, might be correlated with the discharge of specific kinds of nervous energy. However, subsequently developed methods of recording and analyzing nerve potentials failed to reveal any such qualitative diversity. It was possible to demonstrate by other methods refined structural differences among neuron types; however, proof was lacking that the quality of the impulse or its condition was influenced by these differences, which seemed instead to influence the developmental patterning of the neural circuits. Although qualitative variance among nerve energies was never rigidly disproved, the doctrine was generally abandoned in favor of the opposing view, namely, that nerve impulses are essentially homogeneous in quality and are transmitted as "common currency" throughout the nervous system. According to this theory, it is not the quality of the sensory nerve impulses that determines the diverse conscious sensations they produce, but rather the different areas of the brain into which they discharge, and there is some evidence for this view. In one experiment, when an electric stimulus was applied to a given sensory field of the cerebral cortex of a conscious human subject, it produced a sensation of the appropriate modality for that particular locus, that is, a visual sensation from the visual cortex, an auditory sensation from the auditory cortex, and so on. Other experiments revealed slight variations in the size, number, arrangement, and interconnection of the nerve cells, but as far as psychoneural correlations were concerned, the obvious similarities of these sensory fields to each other seemed much more remarkable than any of the minute differences.