I disagree. An excessive emphasis on the one-seat ride results in an unbelievably complex network - in other words, the hole we are in!
The network's complexity has nothing to do with lack of interchanges. It's entirely down to the size, topology and critically the ratio of paths through the hub to total number of stations.
I'm not so sure that boardings (or alightings!) per se are the factor that bleeds capacity and speed. The thing that really causes the bleeding is bidirectional passenger exchange, i.e. when half the full train wants to get off, and when half the full platform wants to get on. This is the killer at Central and Town Hall; note the inaugural deployment of platform marshals on Town Hall #3 during morning peak, selected I believe because it is the finest example of bidirectional passenger exchange. Trains arriving from the Western and Northern Lines disgorge most of their passengers, yet the platform teems with passengers boarding for onward connections to the North Shore Line.
I don't disagree that simultaneous boardings/alightments are both a particular feature and problem with CityRail. However, you have to appreciate that any increases in interchanges:
a) doesn't change this. At best it reduces PAX waiting on platforms by reducing stopping pattern numbers, but not by that much - you're hardly going to get down to a single stopping pattern on Sector 3.
b) re-creates this problem at additional places on the network. (What's likely to happen at Chatswood post the NWRL is probably the best example - it'll be worse the TH).
I agree that passengers dislike this, but I feel compelled to point out that this allows a concentration of resources, in turn allowing higher frequency, which enhances the usefulness of the service.
The same load balancing benefits can be achieved with branching. Compare running a Cumberland Line Richmond-Cambo in leu of Richmond-City. Currently (bar the few peak services), the least used journey that does not pass through the main hub: Main South to Main West. Yet a much more popular journey: Richmond - main hub - has to interchange. And you still need the same capacity (or more actually, because the main south needs more capacity unused west of Granville). You just have more interchanges and associated variance in dwell times & OTR.
Again, I disagree. One need only witness the congestion and waste of the State Transit Authority bus network; as it is designed to give every suburb its one-seat ride to the City, there is the congestion of thousands of bus movements into the City every weekday morning. There is also the much weaker service frequency, as resources must be spread thinner to ensure everyone has 'their' bus, yet due to our geography routes overlap in giant stacks, resulting in parades of half-empty 40-foot buses - waste.
I think you are confusing concepts a bit here. Effective frequency is a trade off between unit size, route demand and your acceptable loading factors. Bigger units means lower frequency for the same loadings. (The SD train argument). With an interchange system the only way you can really increase efficiencies is with larger but less frequent units on the core route. And with busses the efficiency gains of doing this are marginal at best (because there is no such thing as a 400 seat bus). You can also achieve efficiencies with a small with a smaller unit more frequent unit on a feeder route. But in a bus context, you don't save much going to a smaller bus.
But I'll throw you another one: The term I shall summarily invent here is "passenger frequency inconvenience cost". The gross inconvenience factor of the wait based on service frequency. For turn up and go, you're expected wait time it 1/(2f) (f=frequency = 1/t t=time between services). If you have to catch 2 services, a & b, it becomes 1/(2fa) + 1/(2fb) [plus interchange time too], which is *always* larger than 1/(2 Max(fa, fb)) [even ignoring interchange times].
If one of those routes has a non-turn up an go frequency ( lt 10min) - as many of our routes (bus and train) do - you've added the additional complication for the passenger that they cannot turn up at the hub and get on the "right" service, they have to guess when to get on their high frequency service to meet their connection. Where with a timetabled system they could reduce their expected wait times to less than 1/(2f), if there is a high frequency mode they need to allow for the variance in journey time of their high frequency leg to do this. It means their expected wait at the interchange for their least frequent mode is increased, along with the interchange cost. From the PAX perspective, even if the interchange is seamless, there overall journey time is compromised by: 1/(2 x Min(fa, fb)) - 1/(2 x Max(fa, fb)).
By virtue of our geography, most of our collector routes have very small frequencies, and most of high frequency routes still have a high enough run time variance to ensure PAX can no longer use timetables to minimise their wait. And by virtue of poor interchange design (especially involving mode changes) in Australia and NSW in particular, our interchange costs are usually high. On a 15min frequency feeder service, you have all but guaranteed an additional 7.5min (+2.5min interchange cost) increase in journey time. And that extra time is all spent waiting around on a platform somewhere, perhaps "congested", perhaps not. Added to that, 2 trains halves you're chances of OTR (if there is a timetable of course). This as much as anything explains why interchanges are so deeply unpopular in Australia.
The only real solution to this is guarantee connections off specific high frequency services, which completely defeats the purpose.
In the case of the STA, if it really is as you describe (I have no data one way or another), one of the great advantages of bus systems is the unit sizes are so low it is very easy to adjust routes and collection areas to match desired loading factors with desired frequencies. The fact that this may not happen in Sydney is not surprising, but it's nothing to do with passengers wanting direct connections. Direct services exist primarily because it's in the provider's interest to have a common hub they can work out of.
The issue as I see it is you might be increasing the frequency of services on the core trunk routes, but you are not delivering improved frequency benefits to the users, the reverse in fact.
In the context of CityRail, CBD connections are undeniably problematic. However, if the interchanges are made in suburban centres, allowing a more streamlined, efficient and effective service into the City, then surely they reduce congestion.
I think in the particular case of CityRail it's not so much the number of connections but the direction, as per your point. Adding to the boardings at a major destination AM station station like Parramatta or Chatswood is potentially hugely problematic. As is adding boardings at Central.
In the afternoon/evening peak, there is the problem of passengers waiting for 'their' service; much congestion could be relieved by having service patterns such that everyone can board the first service. Currently, at the height of afternoon/evening peak the Western Line has an express and a limited stops pattern to Penrith/Emu Plains, a limited stops pattern to Richmond and a local pattern to Blacktown.
And a main north, and only certain trains stop at Clyde for Carlingford etc

. Plus IIRC some of these trains uses the suburbans west of Strathfield too, meaning they are attracting customers (whoever said I wasn't customer focused?)
But again I think you are mixing concepts here. Fewer simplified stopping patterns vs overall journey time.
And PAX waiting on platforms (so long as they are out of the way) aren't really the biggest problem, it's those waiting to broad specific trains contesting with those wanting to get off.
And even so, are interchanges really an effective solution to the overall perception/reality of "congestion"?
Yet if everyone was able to board standardised expresses and change at Parramatta and/or Seven Hills, everyone would be able to get on the first Western Line service and change later, rather than clog up valuable CBD platform space.
(Yes, there is the problem of Lidcombe, Auburn and Clyde; I have made the assumption that they will be serviced by the South Line. Seven Hills is attractive because it offers cross-platform interchange that neither Parramatta nor Blacktown offer.)
As it happens I agree with you on this point, but not to the extent there should be interchanges. IMHO western line trains should have standardised stopping patterns at least as far west as Parramatta, perhaps Westmead, with the major stations serviced by most trains so the majority of PAX can can get either the first of second train that arrives. I don't see where I'm introducing interchanges here (other than perhaps some non hub inclusive journeys).
To sum up, I think this interchange idea is an ineffective solution to an incorrectly diagnosed problem.
- Observed problem: crowding
- "Obvious" solution for the government: more trains
- Issue: not enough paths for these new trains
- Solution: route them where there are paths.
- Issue: How do the PAX get to where they actually want to go though?
- Solution: "F'n whinging PAX can bloody well interchange like they do everywhere else in the world".