A few more, closer, readings of this book are in order. Thus far I have only really read closely the parts that mattered to me (foundations, function concept, complex function theory, infinite series, algebra leading up to Galois), and perhaps one day I'll be able to get something more valuable out of the calculus and ODE/PDE parts.
Gian Carlo-Rota says that the three volumes (this is just volume 2) are probably the best account of the history of modern mathematics up till the early 20th century in the English language (I have heard that there are even superior accounts published in Russian and German). I am inclined to agree.
The accounts are mathematicobiographical -- there is certainly much said (and much can be inferred) about the mathematicians themselves, but of course there is still a much heavier emphasis on their mathematical legacy. While I have only skimmed through vols 1 and 3, I can say that vol 2 and vol 3 are not very accessible to a general audience (if one wishes to get a lot out of it). Most of volume 1 is easily accessible by a relatively attentive mathematics undergraduate.
On the other hand, vols 2 and 3 probably aren't even accessible to some graduate students, though this is not really due to the difficulty to the material, but more of the scope. Much is said about the calculus of variations and ODE/PDEs in vol 2, as well as Galois theory, elimination theory, and so on. The topics concerned in vol 3 are even more advanced; if "birational geometry" or "ideal theory" sound like gibberish to you, it is likely that little else in vol 3 will be coherent at all.
Of course, the sophistication (compared to its popular counterparts) of this historical account is its primary merit. The bibliography alone would be useful to certain mathematical researchers. The exposition found here still manages to capture the whole essence of the mathematics done by our ancestors, without deforming it all such that it loses all precision and meaning.
Of particular interest to virtually all readers would be a discussion of the metaphysical arguments used by many prominent mathematicians in their work, and a brief (two centuries max?) loss of primacy of rigorous proof in mathematics. Euler himself apparently agreed with the assessment that mathematics would have been exhausted and that in the future, the important problems will only come from the natural sciences.
(The following concerns less about vol 2 but more about the scope of the whole work, especially in regards to certain accusations of the works being too Euro-centric)
Another important note is that the 3 volumes are not meant to be exhaustive at all. The objective of the work is to provide an account on what Kline calls the "main line" of mathematical thought. Other readers have noted that this is too "Euro-centric". Kline perhaps does not help his case by writing:
"To keep the material within bounds I have ignored several civilisations such as the Chinese, Japanese, and Mayan because their work had no material impact on the main line of mathematical thought." [emphasis mine].
Yet any serious reading of the volumes will show that any Euro-centric bias, other than the desire to not create an exhaustive account, is not present. The main line of this mathematical thought is not an imaginary Euro-centric fiction, but an abbreviation of a historical account of interplaying ideas. For example, the mathematics of the Hindus and Arabs, which (deservedly) take up several chapters in volume 1, for their specific contributions directly affected the major mathematicians that worked in the ensuing centuries. On the other hand, results of the Chinese, for example, though some of which were certainly ahead of their time, were virtually, if not completely, unknown to the developers of the main-line of mathematics.
The significance of this main-line of thought is evidenced in volume 3. Two thousand years and 800 pages later, Euclidean geometry continues to haunt the late 19th-early 20th century mathematicians. The main-line is not an arbitrary thread of history, but a coherent, omni-present braid that underlies much, if not all of mathematical thought, up till the present day.
Otherwise, readers still interested in an account of the non-European involvement in the main line of thought need not look towards the past. For if a volume 4 ever to be written (unfortunately, Kline is dead), many contributions from non-European mathematicians will surely be elaborated on. For example, one should expect that any account written in the next 50 years concerning the development of modern algebra and algebraic geometry would expect many chapters dedicated to the works of Japanese mathematicians such as Hironaka, Nagata, Taniyama and Shimura (of Taniyama-Shimura conjecture fame), Nakayama, Mochizuki, etc.