Paul,
I don’t quite agree with all your assessments. It seems to me this conclusion of language simplification is most often reached on the basis of English and Chinese. However, when seen in wider scope, it ceases to hold water.
An infant learns every language with the same ease, an observation on which Noam Chomsky based his influential (though, in some of its aspects, quite ill-advised) theory. One only needs to think of Arabic, which could be labelled as a complex language but has 242 million speakers (according to Ethnologue). The Latin language easily spread (as is well known) to vast areas in Western Europe despite its greater inflectional complexity. This is seen even today in Romance languages. It is not the language but prestige of the culture of its speakers.
Most importantly, however, having little inflection does not mean the language is simple (as testsuda noted). The roots of this notion are in fact in the classification of the 19th century, when (chiefly) German linguists thought that, on the one hand, the absence of intricacy in Chinese meant that the Chinese are simple, and, on the other hand, the polysynthesis of language such as West Greenlandic shows that as it is still undeveloped, so are the Esquimaux. Lo and behold: the ideal state of a language coincided with that of German, the lingua franca of linguists in that period. These kinds of undercurrents have been abandoned, but some of the mindsets persist. I for one do not regard Chinese and English as easy or simple languages, their simplicity being only superficial. As said, this must not be upset with the impression of an infant learning its first language.
Besides, it is easy to imagine English or Chinese (re)developing cases or verbal inflection. This happened for instance in Hungarian, which famously has a myriad of cases. Most of them are relatively late, ensued as originally two words merged into one. Thus we have paradisumben ‘in paradise’ already in Halotti Beszéd, having originally the word bél ‘(the) inside’ plus -n of locative. In Finnish, as well, cases conveying locality are more recent than the so called grammatical cases (like genitive).
As for English, it has indeed been expressed that it would have pidginised due to strong contact with Norman French and especially Old Norse, but this has convincingly been refuted or at least shown as highly unlikely. In addition, it is important to bear in mind that when Shakespeare started his career—that is, when English had already developed more or less to its present stage—English was, besides Britannia, only spoken on the Isle of Man and the eastern and southern coasts of Ireland. The expansive spreading that ensued did little to the language.
testsuda,
As the language learning experience differs immensely between individuals, it is very difficult for me to appraise your ordeals. I can only guess that multitudinous inflectional system of Latin and Ancient Greek may cause you hardships. This could indeed be due to the analytic nature of your native Vietnamese. My mother tongue is Finnish, but as the first foreign languages we learn are generally Swedish and English, everything tends (at least at first) to be viewed through Germanic grammar. Often either languages that are dissimilar to one’s first language or to those that one began with are considered most difficult. It is often said that Finnish is really complex, and Finns like to reiterate it. However, Finnish is not that difficult, only somewhat different from Indo-European languages with which many are familiar.
As to how languages evolve, I think our minds are obfuscated by the at least superficially conflicting knowledge we get from our different experiences. Languages do seem to attain complexity as the tradition of writing them becomes more and more established. I think of Biblical Hebrew here (having highly heterogenous collection of texts) which lacked in some of its syntactic features intricacy which it may have obtained had it had opportunity to develop as a natural language. This is, though, only one way of viewing things.
Literary languages obviously lacks the non-verbal qualities of spoken language and thus has to be more strictly governed. And features like floundering and repetition of content would seem more disturbing in writing than in speech where they may abound.