Grinold Kroner P/E

Change in P/E should be comparison between Year0 and Year1, correct?

And if given, forecasted P/E 5 years from now, then I have to get the annualized change?

Did not found any support or contradiction in text. Therefore, your approach seems reasonable to me.

yep that’s correct. There’s a good question on this in the CFAI 2019 mock