Predicting Release Dates

A lot of discussion regarding new Transformers revolves around "when". As soon as a figure is leaked, teased, or announced, the question is, when will it be available for preorder, when will it be available online, when will it be available in stores, etc. Some fans make logical guesses based on announced release dates, seasonal buying patterns, etc., but it's not an exact science, right?
Strictly speaking, there's too many variables involved in a toy's design, manufacture, shipping, and stocking to make it an exact science, but how would you feel about a little inexact science?
Hypothetically speaking (and I mean that in the non-scientific sense), assuming a person has access to years of news articles recording when given toys were first announced, first previewed, first found, and so forth, a person could possibly use content analysis to quantify the information in those articles, build a database that summarizes details relevant to tracking a toy from announcement to procurement, and use the trends in that data to build a regression model that predicts (inexactly) when new toys will make it to stores based on dates they were announced, teased, etc.
Hypothetically speaking (again, in the non-scientific sense), is that a novelty people would be interested in playing with if someone did all of that? Or does it just sound like a stupid waste of time?
Strictly speaking, there's too many variables involved in a toy's design, manufacture, shipping, and stocking to make it an exact science, but how would you feel about a little inexact science?
Hypothetically speaking (and I mean that in the non-scientific sense), assuming a person has access to years of news articles recording when given toys were first announced, first previewed, first found, and so forth, a person could possibly use content analysis to quantify the information in those articles, build a database that summarizes details relevant to tracking a toy from announcement to procurement, and use the trends in that data to build a regression model that predicts (inexactly) when new toys will make it to stores based on dates they were announced, teased, etc.
Hypothetically speaking (again, in the non-scientific sense), is that a novelty people would be interested in playing with if someone did all of that? Or does it just sound like a stupid waste of time?