The 10% difference sited on old vs new battery may or may not exist. The old battery specs its 9Ah at 0C/32F. The new battery specs its 8.1Ah at 25C. Specmanship in play as the temperature of the measurement is important.
The old battery lists 9AH with a 18A (1C) draw at 0C/32F temperature.
It also has a curve for 20C/68F which shows roughly something like ~7.5Ah-8Ah (hard to get exact with the log graph shown but looks like 24+/-2 minutes until 100% depth of discharge (DoD) is reached).
The new battery lists 8.1Ah with 18A (1C) draw at 25C, about 27 minutes to 100% DoD. The actual discharge graph only shows a 17A discharge curve, but it too looks to be in the 28 minute range to 100% DoD.
http://www.rpci-inc.com/pdf/ES17-12.pdf
I overlaid the discharge curves between old/new. The 17A/18A are green/red curves. What is more interesting are the 36A (old) vs 34A (new) curves. The 36A (old) curve shows 100% DoD at about 16-18 minutes vs 34A (new) curve shows 100% DoD at about 9-10 minutes. The voltage drop under load favors the new battery which maintains a much higher voltage output under similar load. This was confirmed by a quickie test using a 50A dummy load on last year's ES18-12 vs this year's ES17-12. The ES17 had a ~.2-.3v higher voltage while under load. Load was held for 20-30 seconds.
The batteries are close, but are different and diverge at larger discharge rates. The biggest difference seems to be the maximum spec'd discharge; 230A (old) vs 720A (new). Greater safety margin?
Bud