Is there a more concrete way to be able to standardize the GWP-total values?
This is what the impact methods are trying to do. TRACI refers to IPCC 2001 or 2007 (not sure) with the 0/0 approach (no biogenic carbon uptake and no biogenic carbon emissions accounted), CML refers to IPCC 2013 with the 0/0 approach again and PEF (EN15804+A2) refers to the IPCC 2021 with the -1/+1 approach (explicit accounting of biogenic carbon uptake and biogenic carbon emissions, which should result in net zero biogenic carbon impacts, in case a full life cycle is modeled).
The biggest difference of EN15804+A2 compared to all other impact methods is the accounting of biogenic carbon. This will often result in negative GWP-biogenic results, because stand-alone LCA data sets are usually only cradle-to-gate. So if you do not model any end-of-life, you will have biogenic carbon uptake (negative results, because of a positive impact for the environment), but you will not have any biogenic carbon emission back to the atmosphere (which will usually happen at one point in the end-of-life, except you store it magically for thousands of years).
If you subtract completely GWP-biogenic or if you only look at GWP-fossil your values will still not be the same, since TRACI/CML use older IPCC factors and PEF EN15804+A2 uses IPCC 2021.