Susan, Sure - there ought to be a long, slow variance at the scale of the annual seasons. But in principle this also implies variability at other scales - e.g. variability of the success of the hunt in the short term, and variability due to droughts/bounty in the longer. Here a very brief speculation by Nassim Taleb himself earlier this year to an intermittent fasting experiment…
https://drive.google.com/file/d/0B8nhAl ... view?pli=1
…where he calculates that if one experienced a day-long fast once per week, this would also imply a 2 day fast ~1x per month and a full week fast 'every few years.' In principle, 'feasting' would work similarly.
So here a long, slow variability in carbs/fat/protein (and even other micronutrient?) content might also imply other variations on shorter scales - i.e. not a 'carb night' every week either, but ups and downs in carb content on both short and long scales. Don't think we need to over think this as out ancestors sure didn't - they just experienced it based on variability. A confirming clue that it's time for a change for me was yesterday when one of our favorite local farmers told me he would not be there after next week for a while as the fall foods had finished and it was time to replant for spring harvest (I live in coastal Texas).
Hard to know what optimal looks like if true, but seems to me that the most direct implication is that a fairly stable dietary makeup throughout the days, weeks, months and years may not be such a good thing.