There are two established methods for calculating carbon footprints.
The hybrid model uses both providing additional accuracy at the cost of additional work and a bit like a hybrid car has a much better performance.
This required a large upfront investment in understanding the inputs and outputs for the entire global economy. However since economic statistics are a key part of global economics there is a huge effort to maintain and keep these statistics up to date and the underlying data is published each year, which means that they are significantly more current than material and process data.
Initially these were highly flawed due to the ability to set arbitrary boundaries. This has been addressed to a certain extent for materials by saying that all materials need to be measured with a low cut off point, however the standards still miss out services that are part of creating the product and other general overheads.
The next challenge is the underlying material carbon footprints, there are large databases however they cover 1,000 of materials rather than the millions in existence and are often decades out of date.
The EU Product Environmental Footprint Guide requires "fair" quality for 90% of data and and in the final use profile that 70% of specific and generic data be "good quality" with at least 20% "fair" and a maximum of 10% less than "fair" quality.
We have moved from PAS 2050 through ISO 14067:2013 and ISO 14067:2018 to EU Product Environmental Footprints (EU 2021/2279). A significant amount of work has gone into creating Product Carbon Footprints however the number of product carbon footprints is 10's of thousands and there is still the challenge of both limited accuracy and high cost. In spite of the significant efforts to try and create comparable product carbon footprints using life cycle analysis.
In the meantime the NHS has calculated its entire carbon footprint using input output models. The 2004 results were published in 2008 to give an idea of the scale of the task. Co2 analysis then applied Artificial Intelligence to the process and by 2010 had calculated the Carbon Footprint of an entire trust down to individual product level, and in 2010 had over 3.6 million level 0 product carbon footprints.
This allowed rapid carbon footprinting for organisations, based on indentified products that can then have their carbon footprints refined and improved.