As you know, I have been complaining and raising issues regarding the problems the liquidity point system has. Even though my bot (the one with all the 7.77 and 77.7 orders) is actually doing halfway decently and even won an hourly award, I do think the current point system needs to be improved.

https://github.com/steemit/steem/issues/178 has the full discussion, my comment about it is:
"1 minute was too fast, 30 minutes too slow. 30x increase was quite drastic.
Maybe a 5 minute requirement will be enough to create risk for the self-traders, but my bot that doesnt self-trade is getting small fraction of the points that the self-trading bots do.

I know sybil accounts make it near impossible to prevent effective selftrading, but if none of the participants selftraded, I think 5 minute time would work out well. To prevent self-trading some sort of analogue to coinage might be needed. that would avoid the need for path analysis, which can be gamed anyway.

Another approach is to just limit the size of bid/ask that is counted toward the liquidity award. Having 10000 STEEM orders selftrading is really just silly and only being done for the points.

Another approach would be to have a small number of knowledgeable voters to assign percentages of points they should get based on qualitative factors, but of course this is quite messy.

Now I have a halfway decent bot done, I would keep it running as long as I am able to get some sort of subsidy. without any subsidy what would happen is that one day the server would be restarted for maintenance and I just wouldnt find the time to start it up again as the volatility risk does outweigh the gains from spreads from the real trading. As it is the other bots (humans?) tried to fool my bot into lower and lower prices in hopes to drain it of capital but I am tracking implied values from external markets and constraining trades based on that. Though I try to push the price incrementally closer to the theoretical value as that creates the least stress and least risk.

tl:dr 5 minutes and a cap of 100 STEEM size per order is probably worth trying, or have some human involved way to award subsidies. The idea of having a fully computerized metric to award liquidity providers is quite courageous, not sure if it has ever been done before"

I think that having an automated award system is a good goal but due to its large monetary effect it would be much safer to have a portion of it reserved for proven market makers and a portion for the automated points system. When the automated points system is fully debugged, then all the awards can be shifted back to that. In the meantime I am seeing over half a dozen active market makers and with each of them under a performance contract funded by half the current funds would provide more than sufficient real liquidity.

There are 24 hourly awards and I suggest that a dozen of them be reserved for qualifying market makers, 1 award per market maker per day. With about half a dozen active market makers nowadays, that still leaves half a dozen more awards. Now who is able to determine who the qualifying market makers are? It could be done by voting, or selection by knowledgeable representatives, or by the existing group of market makers with a possibility to appeal to a higher authority, or any common sense method. 

The other half can be held in reserve for when the new points system is ready.

Now why would normal people even care about this sort of thing? The reason is that 28800 liquid STEEM per day is 201600 liquid STEEM per week. This is the same amount of liquid STEEM that would come from 104x that, or 20966400 SP being powered down. Even assuming that all the SP is powered down (it isnt), this is around 20%. So in the short term STEEM will get something like 5x the effect of the bitcoin halving as far as supply changes.

This assumes that when bitcoin mining reward went from 25 to 12.5, about 4% of the daily trading volume was reduced. Of course bitcoin trading volume ranges all over the place, so this is a very rough estimate. And this small change in daily supply led to an arguably doubling of the bitcoin price from its earlier price baseline.

You might object that being the same as 20% of SP being powered down is not the same as 20% of daily trading volume. True enough, if we look at the daily trading volume of 2000 BTC, lets say 360,000 STEEM per day, the 28800 is about 8%. That's double the bitcoin halving effect. 

So if the markets understand this, we will see an effect that would be comparable to bitcoin quartering (is that the right word?) or a 75% drop in daily output. This might be too esoteric a factor and we dont really know how much of the daily rewards have been converted to BTC. However, if we see that the price doesnt change much then we can conclude that not much of the liquidity award STEEM has been sold. Keep in mind that if STEEM price doubles to alltime highs after this hardfork, it doesnt prove that the liquidity award STEEM has been sold, I dont think there is away to determine that conclusively.

Anyway, the above is the type of analysis that I believe is necessary to stay on top of the day to day trading of crypto.

I hope this helps you understand the complex system that STEEM is just a little better.

James