I think the fade in time is the same as an LFO's Fade In time. Which scales the output ramping up during the start (1st sequence through however long it takes) and then stays at "100%" after the fade in is completed.

This doesn't delay the movement but if you change MS sequences from one scene to another then the next MS will start and I believe execute the fade-in time for no-sync MS Lanes. This isn't the same as a morph between the current value and a new value since the start of fade in will always start at "0" offset - and this may not match where you were coming from.

A motion sequence is highly leveraged from LFOs so the general concepts would transfer.

Accomplishing what you want inside the keyboard is going to take development of new features. MS isn't going to save the day here. At least not for the most general of cases.

If you want buttons to slowly change Super Knob from one value to another value (and then keep that new value) then this can be done with MS. The problem (and maybe this isn't a problem you'd have) would be to start where superknob is (say in the middle of a ramp). When you have contrived resting values of superknob and assuming you don't move it manually then MS can work. And you wouldn't, by the way, want to use "fade in".

I think the fade in time is the same as an LFO's Fade In time. Which scales the output ramping up during the start (1st sequence through however long it takes) and then stays at "100%" after the fade in is completed.

That suggests to me that the 'Amplitude' clamp value is considered '100% and the delay, either time or steps,
uses either 0 as the low end or uses the leftmost value of step 1 and does a (possibly linear) ramp in the desired amount of time or number of steps.

So depending on the actual shape of the sequence steps it may, or may not, delay much of anything.

[quotePost id=124352]That 'maximum amplitude' is what I don't yet understand. If Sync is Off it could be at any stage of any of the 16 steps.[/quotePost]
You are right, that's why I thought it may 'fade into' the new motion sequence, which would actually be quite awesome in itself. Just as long as the start value is the previous value, and not 0.

[quotePost id=124355]
The problem (and maybe this isn't a problem you'd have) would be to start where superknob is (say in the middle of a ramp).
[/quotePost]

Indeed and equally important you can't transit freely between scenes. With MS you have to follow a predetermined sequence (scene 1 > 2 > 3) but you can't go back or decide to jump between at will (1 > 3 > 2), because your MS start value won't match the end value of the previous scene.

I am fairly certain I could make this happen with external tools (a controller and a script in Bome's Translator Pro), except this would be a common need mostly in live performances. It's a relatively small, but quite impactful feature.

With MS you have to follow a predetermined sequence (scene 1 > 2 > 3) but you can't go back or decide to jump between at will (1 > 3 > 2), because your MS start value won't match the end value of the previous scene.

Exactly. When an MS lane is assigned as a source it ALWAYS provides an offset value when that MS is active.
People keep wanting to use a sequence to do fade ins or fade outs and then wonder why they can't control the volume anymore - it's because the MS is still providing a value. There is no OFF switch for an individual MS.

I am fairly certain I could make this happen with external tools

Possibly. As Jason already pointed out there is no 'current value' setting you can use for a parameter to make the start point equal to whatever the value is at the time you start the MS. But with an external tool you could dynamically modify the MS start value based on the current value of a parameter.