Adaptation vs. aging curves

Back many, many years ago, when I was at the natural peak of long distance running potential, I didn't think much about training and just went out 5-6 mornings a week and ran 6 or so miles at an even easy pace of about 8:00. No speed work, no hill work, no tempo runs, no fartleks, none of that. I ran my first marathon and first 10K (both in Ithaca, NY) during that period. Both the marathon and 10K times (4:14 and 39:36, respectively) are still PRs. Since then, until 4-5 years ago, I continued to run, but at a just-to-stay-fit level (15-25 mi/wk). Since I started to more "seriously" run and race again, I've been chasing those earlier PRs. I'm currently at 4:22 for marathon (Mar. 2013) and 49:35 for 10K (Nov. 2012). I'll be trying to qualify for the Boston Marathon this October (MCM), and the 8-minute drop to a new marathon PR should be no problem. :) But, is a sub-40 10K still doable? Runners my age or older are doing it, so, obviously, it's feasible. But, what do I need to do to train and adapt, to do it myself? Likewise, a sub-3:00 for a marathon is also feasible for my age--and older! (See, e.g., Ed Whitlock and Mike Fremont.) But, how to get there? There's that story of someone asking Jascha Heifetz "Could you tell me how to get to Carnegie Hall?" to which Heifetz replied, "Yes, practice!" What was unstated was that the practice needs to be correct--else one becomes very good in playing badly! Similarly, the answer to how does one get to Boston (or any other time and distance goals in running) is not just training, but the right kind of training. I recently joined the Run Your BQ (Boston qualifying) online community to help guide me towards Boston.

The interplay between adaptation and aging is fascinating! From my time drops in recent races, I know that the adaptation curve is not actually a "curve," but rather a stair-stepped function, driven by a kind of running version of "punctuated equilibria." Does the same apply to the aging curve? Where are you along the adaptation and aging curves?