Hi would like to just give my two bits to help us get a framework around this discussion. It is a really interesting post and could not help reading all the stuff way back from the 06 messages.
We need to look at stocks from four time span perspectives for trends
seasonal, cyclical,generational and eternal
seasonal-these are stocks like sugar and other agricultural commodities which are actually seasonal and vary year by year. These are to be traded
cyclical-stocks like steel companies which go up/down every few years and most commodity companies. These need to be bought on dips, however you never know the final dip

generational-stocks that do well for a extended time period usually a generation when there is a need for the product in the broader economy for many years e.g railways in US for almost 30-40 years when the network was being setup. All commodity stocks for last few years as first China and now India , Indonesia etc go about building the basic blocks of their societies
eternal- haven't found one yet. Buffett claimed that Coke (as a proxy for FMCG)was one such stock , however that is not the case, in the last few years focus has shifted sharply to health based drinks and Pepsico has done much better. Gold is probably one thing which has been eternal
Within these trends one has to find the right stock to ride along and for the right timespan and that is incredibly difficult. Also estimating the timespan and therefore what is long term is next to impossible in most cases
Location of the trends
From time to time within an industry/country we also have micro trends
eg in IT we had desktop computing that gave rise to Microsoftr then offshoring that gave rise to Infosys, then search that gave rise to Google etc
Sometimes demographics can also drive several business trends e.g cars and rock and roll music in post Baby boom USA. Young demographics and coffe shops and cyber cafes in India.
Will the trend make us any money
Sometimes the trend may exist but it will never make you rich. Two examples - Airlines overall, open source software.
Also the trend will have it's hype period e.g the Internet did make money for Google , but several other dotcoms during the hype period failed
how can we optimize our chances in all of these?
Don't know yet but regularly reading stuff and a little luck can raise the chances