It's an unpopular opinion, but I believe in diving standard gases and I don't believe in using a dedicated diluent bottle. This applies mainly to what I do with cave diving, but I dive like that in the ocean. With caves I am rarely at the actual MOD of the gases, and admittedly at 30m in a cave I'm typically on 25/25 because I am very susceptible to CO2 which has a ppO2 a tick under where I normally run at 1.1. That said I have run "trials" where I've run the ppO2 up to 1.6 at depth and then dil flushed and it comes right back to 1.2-1.3 if I'm at the MOD of the mix. It then takes a hot minute with an mCCR to get back down to 1.1, but it gets there.
For me it's all about balances. My sidemount unit really isn't compatible with a standalone diluent bottle that will last for long/deep cave dives which means I have to run my unit off of the diluent bottle/s. On my rack mounted Meg I use the 3l bottle for suit/wing inflation and run the unit off of the LP50's in the rack. If I look at a 30m cave dive with a planned 60minute penetration and have to bailout to EAN32 I am looking at roughly an hour of deco. If I had air in those tanks I am looking at ~90mins of deco. My logic is that if I had to bailout to exit, I really don't want to increase my deco obligation 50%. That comes at the cost of not being able to crash a ppO2 back down in the event of leaking O2, but I can always bailout if the O2 is spiking and then deal with the O2 leaking into the unit while in OC and get back on the loop if I can/go to SCR etc. but if I'm diving a lean dil mix then the SCR deco is even worse than the OC bailout.
For deco, I plug in the offboard deco gases to get the helium out of the loop and it also allows me to do 1.6 ppO2 checks as soon as I get on a deep deco bottle and then let it decay down to 1.1/1.2 then go back up to 1.5-1.6 and let it fall back off. My body feels better with the dynamic ppO2's on deco for whatever reason.