FWIW, I've done research in this space and the definition of a CV event in most trials wouldn't be remotely close to being clinically significant. The one major trial that attempted to answer if revascularization (eg the thing everyone really wants) improves outcomes was stone cold negative (CARP 2008).
CV events can be anything from an asymptomatic EKG change, any chest pain, biomarker elevation, afib, VT, CHF, and death. The really only clinically significant events are afib, CHF, and death, all of which perioperative testing and risk stratification won't capture. Most patients being risk stratified for surgery have enough risk factors for afib. It's rarely an issue pre-op and post-op may need some rate control or a cardioversion, hardly deal breakers. VT is rare enough that you can't risk stratify it. CHF we all try to avoid by resuming diuretics and avoiding excessive fluids. Everything else is clinically meaningless or doesn't improve outcomes (eg abnormal stress for ischemia and not have similar perioperative outcomes).
The original paper was by Goldman in the 70s and it was patients actively infarcting, having uncontrolled arrhythmias, or in active heart failure undergoing surgery. Nowadays, no surgery, unless truly an emergency, would happen under these circumstances. Correct the acute stuff (eg active ACS, control the arrhythmia, get them out of a CHF exacerbation, address severe valvular lesions) otherwise, all that perioperative risk stratification is mental masturbation. You can eyeball most patients and accurately predict their risk.