Well, the upgraded Soyuz would be an improvement, but they'd have done better with some of their more advanced designs that got cancelled, such as the 6-crew Prospective Piloted Transportation System.
There's one interesting capsule configuration that I've never seen proposed, which is to flip and Apollo, CST-100, or Dragon style capsule upside down. That one little flip brings on a bunch of interesting improvements, although the seating needs to re-orient for different flight phases.
With the capsule upside-down on the stack, it allows jettisonable solid abort motors or permanent liquids to occupy the area under the inverted cone, aiming mostly downwards instead of outwards. That eliminates most of the cosine losses of most current abort-rocket configurations.
During an abort, the capsule is already oriented with the heat-shield facing forward, so the capsule doesn't need to do a 180-degree flip prior to either a re-entry or a low-altitude drogue chute deployment.
At allows the capsule to be launched already docked to another module (similar to the Soyuz orbital module) and yet still be on top of the stack for an abort. This allows the capsule to be pre-mated to wildly different types of modules (such as ISS multiple docking adapters, spacelab modules, etc) without worrying about the stress loads added to the capsule, becaus the capsule isn't under the weight of anything above it. This means the abort system doesn't have to be redesigned, upgraded, and requalified by increases to the size of the attached modules (as would be the case with enlarging a Soyuz orbital module). It also means the capsule isn't sandwiched between a service module and a mated habitable module like the Soyuz is, requiring both to be jettisoned prior to re-entry.
This naturally brings about some other spiffy benefits.
Since the stack would have the capsule's heat-shield at one end and a service module engine bell at the other end, docking would have to be done sideways, with a three or four-way docking adapter laid out like a T or + joint. When docking, the pilots would be staring straight out at the side-facing docking adapter through the front Apollo-style windows. The crew could also observe from any attached orbital module. Since the docking occurs sideways, impact shock can be aborbed in bending loads on the entire stack, instead of linearly with the stress concentrated on the crew capsule, which currently is the squashed bug in between a heavy service module and a space station.
In this configuration, another craft can come attach to the first craft's unused docking port, and several could chain together this way. A capsule could undock and leave a fully functioning docking adapter, habitable crew module, and a service module attached to the ISS.
In the case of something like the Dragon's abort system, where SpaceX also plants to use it for a rocket landing, you could even land the capsule upside-down, avoiding the cosine losses and
allowing the pilot a downward view. It would put the landing legs on the opposite end from the heat shield and make their deployment much simpler. It also means the crew could just drop out of the bottom of the craft through its docking hatch instead of climbing out and sliding down or going out a now unnecessary side-hatch. (They could enter on the pad via the side-docking adapter.) For a parachute water landing, the capsule would splash down on the heat shield like normal, and the crew would come out the top.
There are many advantages, but the seats would probably need to re-orient for different flight phases, and that might not be highly space efficient.