In other words, does the growth of total staked PAL also mean a dimished return of a vote per staked PAL?
Exactly this. The reward pool is the same size, no matter how thinly sliced. The more stake actually voting, the less of the total reward pool each controls.
Put another way, the % of the total reward pool controlled has fallen as the % of total PAL base being voted has increased.
The same thing happens on Steem, btw, which is why the returns on voting and delegation are so much higher than inflation. The reward pool controlled by each voting stakeholder is amplified by the non-voting of other stakeholders (like Steemit, for example.)
If I understood you correctly there is an extra nuance I had not previously realised: it's not just a matter of how much PAL is staked, it's more a matter of how much of that staked PAL is used for voting.
In that sense the halving which I mentioned in my post is not only a measure of the spreading of PAL, it is also a measure of how active the community holding staked PAL is in voting and curating. And thus of the succes of PalNET!
A halving in under two weeks then means a huge success. Congratulations!!
That's a fun way of looking at it. I hadn't really thought about that but it's pretty true.
Part of it was also resolution of the decimal precision issues. When smaller accounts voted, no PAL was actually taken from pool because of lack of decimal precision, which left more PAL available in the pool for accounts with high enough stake to actually distribute it.
That's new to me. I thought it was a problem in the GUI. That smaller portions were correctly allocated, just not correctly represented on screen.
Then it's a good thing the decimal precision issue was resolved as quickly s it was.
They were correctly allocated, up until the issuance of rewards on each post. Then decimals were lopped off and only the digit before the decimal mattered (0.999 => 0)