But it then bears the question that if the so called "network effect" companies has any real value (that is to generate profit in the future) or are manifestations of a bubble?
That means they have value for the VCs (usually in the form of "exits", ie, selling to other investors), but the question was whether they'll produce profits (and consequent dividends).
You're mixing up long term probabilities with short term variance and uncertainty. People can make very big bad choices and not see the consequences for years, because not everything unfolds rapidly.
"The market can stay irrational longer than you can stay solvent".