David, Thank you very much for writing this. It is very educational for me to read and I am grateful for this information. It definitely makes the Google APIs a lot more clear. I hope to provide the same benefit to someone else one day.
I thought that perhaps I could have the google API create a spreadsheet with the data and a chart, and return a http redirect linked to the spreadsheet url. But now that I think about it more, this isn't really a good design, even if it is possible: if someone wanted to play around with this and look at the data from a bunch of different angles, they would end up with a whole bunch of spreadsheets they need to delete later. I like the idea of putting the data in the fusion tables, although I will probably have to break it up into multiple tables. The specific data I am working with is SPX volatility data, and I want to be able to plot something like the volatility for the 1 month 25 delta call vs. time. Where it gets complicated is when I want to plot the difference between two time series, such as (1 month 25 delta call - 1 month 25 delta put) vs time. I would rather not create a fusion table for every possible permutation - that would be tedious and would probably put me over the user limit. You say that "tables can be dynamically merged from other tables" - can I do math and subtract one data series from another and plot the result? If not, can I populate two datatables on the client side and have javascript do the math? If I can do either of these things, this project will be a lot easier than I thought. Thank you again, Jim
