Right! There is one across-the-board solution that occurs to me, let's say, the US has an overall cap of 40,000 for employment based preference category, 2800 each are max caps for Indians and Chinese. Because of the 7% quota, I would expect at least some, let's say, M of the 40,000 are left unclaimed. So why can't they take M unfulfilled slots and redistribute that to clear the China/India backlog?

You can think of the same for EB3, and family petitions too. This would be more equitable than 7% quotas.

Edit: I did back of the envelope calculations based on this data: http://www.uscis.gov/tools/reports-studies/immigration-forms... For the last year, there are about 120k applications approved, while 140k is the limit. At the end of the year, the remaining 20k can be split up among people in the queue. This can be made even fairer if you aggregate it over two years, and then tie the number of immigrants to the inflation rate. I estimate it would shave off at least 3-5 years from the current wait times for EB petitions. It can even be tied to the percentage of the country's population that got an H1B. There are some more trends here: https://www.fas.org/sgp/crs/homesec/R42048.pdf