The percentages on the glide path aren't even numbers until the end because each year I'm changing the stock allocation just the right fraction to land at 66% at age 65.
As far as the percentages on the stock sub-allocations, well, they're part "efficient frontier" (William Bernstien), part convenience, and part gut feeling.
Here's a warning about a portfolio like mine - it's a bitch to rebalance if you have multiple accounts - we have two Roths, three rollovers, and a 401k. And it probably won't get much better results than something simpler, like a Vanguard Target Retirement Fund. In fact, it may do worse. But, I just like doing it. Yay spreadsheets.
In an hypothetical efficient market, value and small shouldn't get better returns. But, the data shows that they do, over long periods. Bernstein acknowledges this but I don't know if he's ever come up with an explanation. Maybe modern portfolio theory is wrong and markets aren't efficient.