Likely hypothesis -- three thousand third party application developers who use google web services just got phone calls regarding application speed issues.
Does the report also analyze pure programming jobs, and not team lead, or management roles which previous programmers 'graduated' into, not necessarily became 'unemployed programmers'?