A few thoughts/observations:
In the long run, and on average, a 100% stock allocation gives the greatest return because stocks grow more than other asset classes. So if your goal is to grow as much as possible, you should be 100% stocks. There are two problems with this approach that call for some other asset class.
Since many of us don't have the psychological aptitude to ride out the market downturn without selling, we manage that risk (of selling low) by having an asset allocation that is not 100% stocks. Now, in the face of portfolio decline we have something to do (rebalance) and see that our decline is not as bad as the stock market alone. The act of rebalancing is to buy low and sell high (relative to the last rebalance) and we rightfully feel good about this decision.
The other problem with 100% stock allocation, is when if early retirement years coincide with a market downturn, and we're forced to sell depreciated stocks for essential spending. Having some other asset that hasn't depreciated (at least not as much as stocks) would be better to sell because the loss would be less.
Then as we age, and our nest egg gets spent away, a decline in stock price on a high stock allocation has a larger effect on portfolio value. By increasing the allocation to bonds, later in life, a market price drop in stocks has less impact on portfolio value, hence less chance of running out of money. Of course, we give up the chance of large gains for this stability.
Given all this, it might seem that getting AA right is really important. But it turns out that the particular number doesn't matter much, nor does the frequency of rebalancing. But for the reasons above, I think it is really important to have an AA number and stick to it (or at least stick to a glide path). Failing to do so just means you are a market timer.... Which is fine if that's your thing (its just not my thing).
Here is an article saying the number doesn't matter much
https://www.aacalc.com/docs/aa_sensitivity This calculator takes a much broader approach than many by including SS and pension as part of the "bond" portfolio. Worth a read if you haven't considered that (click on the "Home" at the top)...
For rebalancing frequency Kitces talks like it's important, but the first table in his article shows that it doesn't matter much:
https://www.kitces.com/blog/best-op...y-time-horizons-vs-tolerance-band-thresholds/