I would record the cookie consent choice beside each lab run too. Comparing an accepted session with a rejected one could hide the cost of those new marketing scripts, even when the URL and device stay the same
Consent beside each lab run is the right companion to URL and device. We would keep the weekly retainer series on one mode — usually accepted, since that matches most real sessions — and only add a rejected pair when you are costing the marketing tags. Mixed modes week to week look like regressions that never shipped.