Read Jason's thread 'how to see hidden replies'.

I am fairly new to sound design

LOL! Everyone is new until they do it for the first time.

The sound design method that works best for me is to build from the ground up. Create the basic, ground-level pieces first and then combine them.

The first suggestion I make to people wanting to design sounds, and the one usually ignored the most often, is to keep their initial focus on the word DESIGN. 

The first step in  building anything is to create identify the requirements. If you want to build a house you don't just start throwing some lumber together you figure out just what you need: kitchen?, how many bedrooms?, how many bathrooms, etc.

What you provided is a pretty good start so lets expand on that.

1. you need two layers for now - we'll focus on that

2. how do they need to work together? You said start with pad and gradually bring in a synth lead. When you bring in the lead are you keeping the pad? or are you fading/cutting it out and doing solo with the lead?

3. build each layer separately - you can build them in the same performance but if you do you should MUTE one of them while you work on the other. One thing at a time.

4. get the first part working the way you need it. Will it work as a standalone part? If so make sure your volume, mix and effects are appropriate for that.

5. does the first part also need to work in conjunction with a second part? You said yes - but if keep the first part AND add a second part some things may need to be adjusted. Volume - for instance. The first part volume may need to be changed to work well with your second part. If so you need to figure out just how it needs to change. What other changes need to be made to the first part? Effects?  Identify things that need to change. 

6. still working with the first part - work out how you are going to make the needed changes to the first part. Will they be gradual? Or, perhaps with effects, just lop them off? How are you going to make that happen? Scene change? controller? performance change?

7. the goal is to have a part one that works properly by itself AND that you can manipulate (scene, controller, performance switch) to work with a second part.

8. once the above is complete for part one put part one aside (MUTE it you started with part two in the same performance) and repeat the above for part two.

9. Once both parts have been prepared individually, without ANY performance level settings involved, you can work on combining them.

I usually just create a new one part performance for my part 1 until I get those steps done and then create another new, one parter to get the second part up to speed. For me, that eliminates the impulse to start combining the parts before they are done.  

While it works in theory; I am running into issues with abrupt jumps in volume and timbre as I transition between the layers. Is there a way to make these transitions more gradual and natural?

A common cause of that is that the default volume setting for the performance, or a part, often needs to be different if you are playing the part by itself or playing in along with a new part you 'gradually' bring in. When you play manually you need to have volume set to a real value - zero means you won't hear anything. But when you use a controller to control volume you sometimes need to have the perf/part volume set to 0 since the controller only applies OFFSETs. If the volume is set to 70 and your controller curve can only increase the volume then the volume will NEVER go below 70.

So the curve used by the controller can cause 'jumps' when you bring a second part in.

My advice, again, is to work with ONE part at a time and 'sound check' your performance/playing to make sure that sound does what you want it to do as you play and manipulate any controllers you have set up or change any scenes you are using. Check both sounds individually.

Balancing the EQ between layers so they dont muddy each other up. Setting up effects that can be applied selectively to each layer.

EQ, panning, effect usage and volume all need to be planned in advance and designed. One part at a time - what EQ does it need when played alone? does it need different EQ when played with a second part? 

Solo piano often needs different panning than piano played with other instruments. Reverb is often different for solo piano also. 

If EQ (or others) for part 1 needs to change when you bring in part 2 you need to figure out a way to make it change. Does it need to change gradually? Or can it change abruptly? Abrupt changes can be made by duplicating part 1 and using different EQ/pan/effects for the duplicate part. Then you can use scene changes to switch from the original part 1 to the duplicate with the different effects.

I have tried setting up two different layers within the same patch and assigning each to separate parts in the performance.

I don't understand what that means. How do you assign a layer to a performance. What do you mean by 'patch'?

I'm curious why you are copying a part to have two of the same Parts.  Sometimes there are reasons to do this, but I just want want to understand where you're coming from.  Is the synth lead and pad just tweaked versions of the same sound and you need both of these sounds - built with the same base materials - to be heard at the same time?  Any other reason for this?

I didn't read the linked thread but the title seemed focus on scenes which need to be crafted carefully not to interfere with the superknob if that's what you're using to "morph" between the pad and lead as the dominant sound.   What I mean specifically is that scenes can turn on and off each section (tab) so pressing a particular scene button doesn't mess with settings that are listed under that tab.  There's a superknob tab and I would turn that off for all scenes if you use scenes so the scene doesn't thrash your superknob.   I'd want the superknob to "always" be under manual control and leave the scenes for instantaneous changes.   The on/off for scene sections is termed "scene memory".

I wouldn't worry about EQ or effects until you get volume alone working like you want it.  

CFX+FM EP is the first Performance that morphs between two sounds.  Increasing volume of the FM EP and decreasing the volume of the sampled (CFX) acoustic piano when you spin Super Knob one way and vice versa when spinning the other way.  I can't remember if it's a 2 part Performance or not which would make it easier to look at if it were.   If not, there's likely a better reference for morphing that's simplified.  Check out CFX+FM EP if you're curious.   I'm thinking the lead and pad is trying to work sort of like this morphing example.   Maybe not completely getting rid of the pad.  Maybe so.  It's easy to modify these aspects.