this is just one person's opinion, but maybe it will help you understand why
latex is the way it is.
latex is a "front end" to tex. tex is a macro language, so "problems" (think
about them instead as "how do we get the results on a page looking the way
we want them to?") are inevitably going to be "solved" with macros. the tex
language was created by one individual, originally for his own use, and
subsequently given for free to the rest of the world in an amazingly well
documented form, with a promise to fix bugs (with a reward to the first
finder) that are found in the code or the documentation. (and it may be
that more bugs have been found in compilers by applying them to compiling
tex than have been found in tex itself.) after letting tex loose, its
creator returned full time to his intended life's work, completion of
his magnum opus the art of computer programming.
what tex is not is a large piece of software produced and maintained by
a large corporation. it is a stable platform available for use by anyone
who finds a need for what it offers.
latex was also originally created by one individual, for his own use. it
found favor among computer scientists, mathematicians, and other authors
in the "hard" sciences who found that it produced results immeasurably more
attractive (and easier to read) than what could be produced with a typewriter
or line printer. the original author of latex has also gone on to other
pursuits, after ceding the authority to make updates to a (well qualified)
team of volunteers.
with the exception of a few small commercial enterprises, nearly all
development of packages for tex or latex has been done by volunteers, often
working alone. there are groups of like-minded users who have banded
together into formal user groups, all of which (to my knowledge) have been
recognized as tax-exempt charities in their home countries. there is no
"central authority" that decrees how something should be done. (there are
a number of publishers who depend on (la)tex to produce their publications,
and all of these have their own rules, but there is no coordination among
packages for latex are, for the most part, created by individuals who have
particular needs, and, for the most part, "donated" to the community by
posting on ctan (another volunteer enterprise). many, if not most, package
authors welcome feedback and bug reports, but writing packages is not their
key occupation, and life has a habit of getting in the way.
so, back to your comments. yes, latex is fragmented, and isn't likely to
become any more unified, at least not in my lifetime. all that can be hoped
for is that package authors pay more attention to "best practices", current
conventions (which do change over time), and the quality of their user
documentation. they can also communicate with other package authors and
users (good places are through the above-mentioned user groups, and forums
such as this one) to try to approach "problems" in a consistent manner. but
individuals are individuals, and "controlling" them is somewhat like herding cats.
as for "the best way" to learn latex, i recommend starting with a good book.
my favorites are kopka & daly, and gratzer's "more math into latex". for
special situations, the several "companions" are solid, if likely to become
dated. start with the basic document classes --
proc -- and explore other approaches as well: the ams classes, classes
memoir; all have valuable, and different, points
of view. ask questions here when you run into a thorny patch, but be sure
to say what you've done and provide an example that demonstrates the problem
to give a helper a solid place to start debugging.
and don't give up.