I think it will come back some point in the future for this reason.
there's no point in reducing the minimum fee as long as blocks are full.
I see a time coming where blocks are very rarely full because the cost of on-chain transactions (in fiat terms) has become so high as to push the majority to only use off-chain transactions.
Interesting! Considering the current consolidation happens at 3 sat/byte, you could probably consolidate at least everything between 20,000 and 50,000 sats at 1 sat/byte. Make the transactions separate from the higher amounts, and just keep broadcasting until it confirms.
We use a formula to calculate the fee and a percentage of the value of the input. So when the network is busy like it is now and we are sweeping at higher than 1 Sat/byte only higher value deposits are swept. eg
https://btc.com/05255aa61e2ffeebdab7cd3b0899250c4f133a93f09b0b4bbdd0c67a6b981a3750,000 Sats is the limit for a fee of 1 Sat/byte.
I checked
all addresses with less than 50,000 sat. The total balance is 18,088.92659937
BTC. That means FreeBitco.in owns 1.266% of all dust.
In total, 22,918,211 addresses hold less than 50,000 sat. Assuming an even distribution, 1.226% puts you at around 290,000 inputs. This is enough to fill about 43 blocks. Assuming my estimate on the number of addresses is correct: consolidating everything at 1 sat/byte would cost 0.43
BTC, which is only 0.19% of the 229
BTC value.
I'll have to check with wetsuit but I think the even distribution isn't that even. He said the cost would be greater than 5%.
And of course implement Segwit for future deposits

That is something I have been keen to do for a long time. The problem is that we have always said all your old deposit addresses will remain valid. We would have to run a new system alongside the old one and thus double our overheads.