Wow that is quite the analysis... if elevators are so much better then why dont they use them in underground parking garages today? How do you resolve the issue of volume. If thousands of cars need to get into the tunnel at a certain point at the same time. Sleds seem crazy for the same reason. You will need thousands of them and they add a lot of complexity. I would say sleds would only be used for ICEv and a toll would be charged to use them, otherwise EVs would use FSD and receive signals from the tunnel's brain for control and navigation. I could be eating some crow, but it seems over complicated and thus expensive and prone for technical issues.
One more for the list:
If the system is 3-D, elevators allow stopping at the appropriate level for the tunnel you need. (Although, for single tunnels less than 40 feet deep, the elevator can be hydraulic which reduces complexity and parts.)
An underground garage wouldn't gain much from the use of elevators. The structure exists as a surface for cars to sit upon. The asynchronous/ random access entry/exit timing of the vehicle requires a travel lane along with the parking areas. So, given a travel lane must exist, and the surface for parking must exist, the most efficient configuration is to make the parking and travel surface also the ramp to other levels. Due to the necessity of the travel lane, the ramp is not pure wasted space.
For the Boring system, the access is synchronous/ FIFO, and the purpose is to get vehicles from a single point to one (or one of many) tunnel access points, so the less resources that takes, the better.
The sled gives the system a higher level of reliability/ stability by eliminating the variables inherent in using the traveler's vehicle.
Problems that exist when using vehicles as the motive means:
-Low battery
-Low tire pressure/ flat during transit
-Human override (for vehicles with internal controls)
-Communication with the vehicle in a real time setup (what generation can travel)
-Braking authority
The sleds would have a regular maintenance schedule, self diagnostics, and the ability to charge from the tunnel system. They could also have train style couplers to maximize density.
Real life anecdotes : I used to commute on I-275 from Canton, MI. First exit south of Ford Road is Michigan Avenue. I learned quickly to get the left lane due to the off ramp backing up onto the expressway every day. Not optimal.
20ish years ago I interviewed out in Portland. They had the red light green light set up to stagger entrance to the expressway from the on ramp, kept traffic flowing.
A ramp system has a higher rate it could get cars to the tunnel, but that capacity is overkill. If the tunnel is one lane, then the number of ramps to fully saturate the tunnel is 120 MPH/ avg ramp speed. So if you have 30 MPH zones feeding the tunnel, the 5th ramp is going nowhere. Likewise, in the unload direction, if > 25% of cars are going to a single ramp, the system, backs up. With elevators, the individual rate is lower, but the aggregate rate spreads out the load.
Let's say the tunnel goal speed is 120 MPH and the sled spacing is 140 ft (18.5 ft + average 0-60 braking distance).
Per mile of tunnel we can fit (5280/140) 38 vehicles maximum with 76 vehicles passing a point in one minute (4,560 per hour).
The time to intake a car if the start of the tunnel is open is 140ft/(120MPH*5280ft/mi/60min/hr/60sec/min) = 140/176 = 0.79 seconds, or 1.25 cars per second.
Now this is the maximum intake rate assuming an empty tunnel. An average commute in LA is 8.8 miles per Marketwatch 2015. So lets assume a vehicle trip distance of 8.8 miles (one off one on).
With uniform distribution, each mile loads/ unloads 1/8.8th of capacity. So for each mile, you can exchange 1.25/8.8 = 0.142 cars per second or 8.6 cars per minute. If the cycle time of the elevator system is 1 minute, then the total number of elevators needed to saturate the tunnel is 8.6 elevators per mile or one every (5280/.6) 614ft, roughly every two blocks.
Where the system works well is having multiple distributed elevators to spread the load/ unload. With the elevators/ sleds, the system knows where there are open arrival spots to place you. With computer planning you are delivered, if not exactly at your destination, at least close to it and likely pointed in the right direction. There will probably still be queues at the unload points to buffer for the elevator since the humans at surface level are less deterministic regarding getting on/off the elevator. It is also more fault tolerant in that the loss of one elevator has less impact than a traffic jam/ accident at a ramp.