Lets put it this way... As long as you are handling the 3D art part, you should concentrate on how you can generate enough assets for the art style you are going in a sensible amount of time.
That might mean you need to adjust your art style to one that is easier to model for, or adjust your game design to one that demands less assets or simpler assets.
For example, if you go with a non-realistic style you can skip sculpting high-poly assets. You might still need to retopo your models after finishing the raw models, to reduce the amount of polygons some, but it is the very fine details that take the longest to sculpt. If you can create the game ready lowpoly models without the need of a retopo, you will save time again.
Which is a case of choosing an art style that allows for simpler 3D modelling.
Another thing is, if you choose a game design that avoids organical creatures, you can skip the whole rigging / weightpainting part. As every separate part is "rigid", animation can be done simply from script. In Unity you can actually just load a multipart mesh, and animate the model by moving the submeshes (which get imported as separate game objects, at least for an .obj file).
This can save you a lot of work, though you need to be aware that it is horribly inefficient for the renderer to have models consisting of seperate submeshes, as each mesh gets drawn in its own draw call (clogging up the CPU, and creating a bottleneck).
If you can spend 45$, there is a thirdparty asset from the store, which can bake a skinned mesh directly from such a multipart model in the Unity Editor, the Meshbaker. Saved me a ton of drawcalls without needing me to go into Blender and do this all by hand (which is also possible, have to do that now that I switched to Unreal Engine).
Point is, you save the time needed for weightpainting your models for animation, setting up a complex rig and recording all the animations by going with a "vehicle only" or similar design. Which is an example how the game design you choose can influence the complexity of models.
Never, ever feel that getting third party assets is cheating or cheap! Byuing stock art or downloading free models can save you a ton of time on models that aren't so important (for example environment models, rocks, tress and such), freeing your time to be spent on the more important stuff (players characters, enemies).
Also, you can get many good and ready made parts like textures and stuff for free or at least cheaply if you look around online.
Last point about 3D art is that it is really an "art"... as in takes a lot of time to master. As long as you have fun with it, just continue doing it, have fun, create awesome 3d models, and get better at it over time. In a year what seems hard to create today and takes a lot of time might seem easy to do then.
Now, as for the "biting off more than you can chew" part... yes, most probably you are. If you both are new at it, you will most probably need to rescope to get faster results, or you need to readjust your milestones to a more realistic timeframe.
Learning the needed skills to create any more complex game takes years... creating such a complex game by yourself or with two people also takes years.
Another fact is, your first few projects will most probably fail because of lacking skills or wrong scoping. While working in a team can be good, even during early stages (many peope feel more motivated by having some kind of "obligation" to continue working on the project because other people are also involved), not many teams will survive those early years filled with newbie mistakes. You and your friend need to be aware of that, and be okay with it... like, you might make a game design "mistake", your friend migth program a feature that cannot be used in your game. The feature needs to be thrown away (or at least put on the backburner for now)... it is important that your friend can let go even though he might have invested a lot of thought and time into the feature, and he shouldn't blame you (because neither of you both could have known better in advance, you both are still learning, and making mistakes is how most people learn).
So while working in a team will become a big boon later on, in the first years of learning the ropes a team might hinder you, because you get afraid of making mistakes.
On the other hand, as long as you two can sort it out, being able to concentrate on learning one thing without having to worry about the other (programming for him, 3D art for you) can save some time, and allow you to get better at your specilization of choice quicker.
Most people would now tell you to go back and start with pong, pac-man and tetris before trying your hand at complex 3D games. While that certainly is a sane choice, I'd say you just need to be aware what you are in for.
If you both set realistic timelines (game might not be in a playable state for years to come), start with you dream game today. Small adjustments might make the goal much more approachable (like leaving off organic creatures for now, settling for a cartoon artstyle, or choosing a viewpoint which doesn't allow to see your meshes from close).
Just do not expect to finish such a lofty goal within months when you both clearly lack the skills for that.
