Bad Math Leads to Exaggerated Headlines

Over 90% of US Population Growth Since 2020 Came from Hispanics
https://www.bloomberg.com/news/articles/2024-08-07/over-90-of-us-population-growth-since-2020-came-from-hispanics
 
Census shows America’s post-2020 population is driven by diversity especially among the young
https://www.brookings.edu/articles/census-shows-americas-post-2020-population-is-driven-by-diversity-especially-among-the-young/
The Hispanic population contributed the most to these gains, accounting for 3.2 million, or 91% of the total U.S. population gain of 3.4 million for this three-plus year period. Within this period, the peak pandemic year of 2020–2021 showed smaller annual gains (or greater declines) for each group, but the gains picked up through 2022–2023.

 
What is wrong with the above noted (Bloomberg) headline and (Brookings) statement?
Following the logic of the above statement (which uses net population gain in the denominator), we will have the following nonsensical figures:

• Hispanic population contributed to 91% of the total net U.S. population gain
• Asian population contributed to 34.3% of the total net U.S. population gain
• Mixed race population contributed to 17.6% of the total net U.S. population gain
• Black population contributed to 17.4% of the total net U.S. population gain

 
It would be more accurate to state that, amongst the groups seeing population gains, Hispanics accounted for 56.7% (3,155,631/5,568,319) of the gains and Asians accounted for 21.25% (1,183,045/5,568,319) of the gains. 

Note: 605,849+22,854+600,940+1,183,045+3,155,631=5,568,319