Iso cabs generally leak some sound, unless they're seriously huge and heavy. I built one a few years ago to hold a Celestion 10" Greenback, which is a pretty insensitive speaker as guitar speakers go. The inner and outer boxes were 1/2" MDF with 3" of heavy rockwool and 1" of the acoustic foam used to isolate drum kits from stages between them and under the base and it was good for around 30dB of noise reduction. This was not a small or lightweight box.

Which meant leakage of around 60-70 db with it on a concrete floor, and noticable added rumble if it was put on floorboards. It's surprising how much noise even a 15 watt amp can push out. Still, getting down to the volume of a moderately loud conversation was an improvement on what went before.

These days I often use a Palmer speaker sim DI box and a Palmer reactive load to replace the actual physical speaker and provide the required load to the amps. Works pretty well, I can't easily tell the Palmer (via nearfields) and a guitar speaker apart after a minimal bit of eqing, at least not until pushing the kind of volume that makes a speaker really start to break up. It might not be the exact response of a speaker, but it's very close and does away with mic placement, room issues and mic frequency response issues completely. I've never quite understood why so many emulators are big on emulating microphones - using a mic is a necessary compromise with a real speaker, but it is always a compromise and to my mind one that would be better done away with if possible.

Tlw, don't they reverse recordings made with Mics to try to back into the emulations?

I think that is why they emulate Mics.

Pragi
Chuck, 
hope you are getting your Iso cab as silent as you need it.
Have you tried to use ?
1.Inside 10 cm thick acoustic foam  and
2. to fill  the joints with silicon and the lid with sealing tape ?
 3. 3 close fitting frames for the lid
Off course, my silent box is  not totally silent but I can 
play very  loud  during the day time in our house.
 
During playing very loud via silent box it is possible to speak  quiet
in the same room.
 
Hope this  is anyway helpful for you.
regards
 
 

Those are some great ideas Pragi. I thank you for that.
I think caulking the joints with silicone is my next step.
didn't use rock wool but I probably should have. If this next phase fails, Im just going to make a 3 inch form in the corner of my basement and make it out of concrete .
 
Seriously, I swear I've thought about it.. LOL.
It wouldn't be very mobile but im absolutely positive I wouldn't hear it.

Wouldn´t use rockwool cause of its effects on the lungs.
10 cm (3 inch?) thick acoustic foam appears to be in my box 
very isolating but a bit more expensive,
Imo basotect is a good choice as  rockwool replacement
To reduce the vibrating and resonating of the isocab replacment feets for washing machines are
a cheao option..
regards

a double

gswitz
Tlw, don't they reverse recordings made with Mics to try to back into the emulations?

I think that is why they emulate Mics.

No idea.

I would imagine if speakers were recorded to get frequency and impulse response plots to use as the emulation model then a flat measurement mic in an anechoic chamber would be the way to go, then recurve the emulation plot to remove any microphone related effects, but I could be wrong.

The whole business of emulating microphones leaves me scratching my head to be honest. If e.g. a vocal is recorded using, say, an SM57 then imposing on top of that recording the frequency response and distortion characteristics of e.g. a vintage Neumann strikes me as the same sort of thing as using a Line6 Pod into a guitar amp without switching the Pod's speaker emulation off.