I do not wish to preclude the slow supply when demand is low, I have always valued it.

The block maturity would remain at 720 blocks, going through two diff changes rather than one.
The diff could still change +/4x over 720 block periods.
The proposed change is nothing to do with miner profitability, it's entirely to enable the network and market to find an equilibrium that is fair in the future.

We currently have two difficulties, 500 and 2000
this corresponds to 17.5 GH/s and 70 GH/s network speeds
which corresponds to roughly 4.35 BTC to 17.5 BTC per day production cost
which corresponds to roughly 30k satoshi to 120k satoshi market price if everything is sold
take in to account miners holding some, and you get 43k to 175k satoshi market price.

Our current status is:
that if the market price is above 43k on low difficulty, the difficulty will rise back up 4x and effectively stop the supply
or if the  market price is below 43k on low difficulty, the supply will start again at 14,400 btm created per day.

Now, what has happened, and will happen again, is that price will go above 43k in high diff period (due to constrained supply), and force the 4x rise in difficulty back up, and stall the supply again.

The only way we can break this cycle without changing the constraints of the diff algo, is for one of the following to happen:

1. keep the market below 43k satoshi during high and low difficulty, then provide enough market liquidity to buy 10k+ BTM per day
... this would require about 15 BTC volume daily minimum and 4BTC of that purely to swallow new coins. The market price and diff would likely both keep dropping until high demand nudged it up a little bit.

2. get the market to be stable at about 175k satoshi, which would require daily volume of about 70 BTC, and about 17 BTC purely to buy newly mined coins. Again the market price and diff would likely both keep dropping until high demand nudged it up a little bit.

Due to the cycle, neither of these will happen under any circumstance, during the high diff period and given rarity, it only takes a small amount of BTC orders in that full time (now 3 months) to push the market price in to the 43-175k range.  Look at the market you'll see it now, this rise has ensured we're about to enter a high diff straight after the low diff.

Why does the high diff happen? Mining profitability calculators, when low diff occurs mining cost will be about 30k satoshi, market price will be about 60k, so miners will see "200% profitability" and jump on network with 100GH/s upwards, pushing the diff back up to 2000 within hours.

So what does the proposed change do?
The proposed change allows the difficulty to change half way through a cycle, by half the max amount. Diff 1000. Practically this means that mining profitability calculators will show 80-120% profitability after 360 low diff blocks. This should have the effect of some network jumping miners switching off, and bring us to a difficulty of about 1500 for 360 blocks.

When diff hits about 1500, we'll enter a new slow production phase, which won't be quite as extreme or quite as long, maybe 1 month with 10-20 blocks per day, followed by a low diff at 750.
If that subsequent low diff at 750 equates to a product cost below market price we'll hit high diff again and slow supply, if not diff will lower further to meet the market. The same pattern as now but more constrained, and with the opportunity to find equilibrium.

Supply and demand will not be changed by this, nor will mining profitability, if marking isn't out we'll have a high diff and low supply, if it is out we'll have a medium diff and started supply.

More likely is that it will allow this extreme high low cycle to reign itself in over say 6 months to equilibrium, with supply gently increasing as demand grows, and knock on of network slowly becoming more usable for transactions as that demand raises.  Which is what we want, and what is currently precluded by our extreme cycle.