Public Transport has two primary goals, which actually compete with each other rather than complement each other - Patronage and Coverage. This distinction between the two functions matters a lot for how public transport networks are designed, and which projects are worthy of funding - but it's a distinction that's rarely made explicit. So you'll often have one group of people arguing for a project on the basis of Coverage and the other against it on the basis of Patronage - neither side is really speaking the same language, so the debate goes nowhere.
The goal of Patronage aims to place your train lines or bus routes where they will attract the most passengers - so it needs to go from where people are, to where they want to be, as quickly and efficiently as possible. Patronage-centric systems are often the best bang-for-buck in terms of taxpayer dollars (or private companies' profitability). Also, when you get to the kinds of densities you see in modern cities, it's often the best or even only way to move those kinds of volumes of people around. Patronage-based routes will be fast, direct, and carry huge numbers of people. Patronage goals see public transport as a better way for car-owners (or people who could choose to own a car) to make certain trips. A Patronage-based public transport system can freely compete with the car for mode share.
The goal of Coverage aims to provide some basic level of service to the maximum number of people possible. This is broadly interpreted as having the most houses within walking distance of a stop as possible, so when you look at a map of the area you can say "Yep, most of this suburb is covered by public transport." It's treated as a lifeline - it has to take a detour to go past the house of the old couple who has been riding the bus for decades, because that couple has no other means of transportation. It will meander around between several of these detours, and eventually get somewhere worth going (a train station, a shopping centre, etc) but it will be a lot slower than a car. In addition to being slower, it is likely to be very infrequent (ie it won't run every 10 minutes, even in the peaks - it'll be an hourly or half-hourly service) because otherwise they would be totally empty, instead of just mostly empty. At the extreme end of the Coverage scale, they are usually so inconvenient that anyone with a viable alternative will take it,so you only have people who absolutely cannot drive, cycle or walk. The Coverage goal is no less important than the Patronage goal, because services like this really are essential for many members of the community who have no other means of transport.
Most public transport networks have to be a tradeoff between these two goals. As soon as you start making detours here and there, on narrow streets with low speed limits and speed bumps, you start losing the patronage of people who can just drive along the quickest route; as soon as you limit the number of places your train stops, you limit the trips that are possible with it and therefore limit your number of passengers. Any service that puts too much emphasis on Coverage will become so useless it gets no patronage, and any service that puts too much emphasis on Patronage will have no coverage and therefore become useless to much of its potential market.
In practice, different routes within each city will lean towards one goal or another to varying degrees, and this is often the best compromise. Speaking very broadly, trains are best-suited to Patronage-based routes, and buses are best-suited to meandering Coverage-based routes. So your trains are your trunk lines, that carry huge numbers of people, and go fast but stop large distances apart; and your feeder buses bringing smaller numbers of people in from lower-density areas, so they all converge on the train station. In Melbourne, trams fall somewhere in the middle - they tend to be fairly direct but they also stop frequently and mostly have to share road space with other traffic, which slows them down.
It is important to remember, though, that buses are capable of serving Patronage goals if you set them up properly. This is particularly relevant for regional cities that just aren't big enough to support a rail network - if everyone from X outer suburb travels down the same arterial road to the centre of their regional city between 8-9am, then a frequent bus with a straighter path and wider-spaced stops can serve a Patronage goal much more cheaply than constructing train or tram tracks. The road is already there, and if it reduces car traffic it won't even add to the road maintenance costs. Each transportation mode plays to its strengths, and it's about picking the right mode and right route geometry for the task.
The problem often comes when you start trying to mix and match the two goals within the same project, or using the wrong mode to achieve a particular goal. The example that's being debated at the moment is the possibility of returning passenger trains between Geelong, Ballarat and Bendigo (which is a regional inter-city system, but I don't really have an intra-city example on hand). There are already buses that travel these routes and stop at various small towns along the way, but they are slow, uncomfortable, and stop frequently. They also run infrequently, so you really have to plan ahead when you'll be taking it, and hope that there happens to be a bus around the time of day you want to travel (unlikely, with 3 per day). This means they serve the Coverage goal well - it may not be too convenient, but it is at least possible for a person in Meredith to catch the bus to Ballarat or Geelong. So a teenager without a car, or an elderly person without a car, at least has that option.
But it also means they serve the Patronage goal poorly - it is slow and uncomfortable. The hundreds of people who drive along these roads every day are not going to be tempted to jump on the bus, and nor are Ballaratians likely to say "You know what? I think I'll jump on the bus to Geelong." on a whim. More efficient movement of existing travellers, and induced travel - these are the outcomes you're after when you want to introduce a train, and this does not happen with a bus service like this.
All versions of the plan (from different political parties and lobby groups) involve stopping at the various towns in between the big regional cities. This, to my mind, is a result of blending the Patronage and Coverage goals in a way that simply doesn't work.
From a patronage point of view, it should be a train that goes from Ballarat to Geelong and vice versa, stopping only at Bannockburn on the way. Bannockburn functions as a dormitory suburb of Geelong - lots of people choose to live there for its small-town vibe, but travel to Geelong to work or study. Bring back the train, and people who currently drive will switch to the train; furthermore, people who want to work or study in Geelong will suddenly consider moving to Bannockburn, because with the train it suddenly becomes a viable option to live there.
Bannockburn is important, but the real drivers of this will be people going from Geelong to Ballarat and vice versa; so your service must be set up to be a viable alternative to driving for those people. I think it's reasonable to bring the pendulum a bit back towards the Coverage end of the spectrum, though, so stopping at Meredith - the biggest and most central of the remaining towns on the route - is probably a good idea too, because going all 65km between Bannockburn and Ballarat without stopping does limit your Coverage too much. But stopping at every tiny hamlet on the way drives up the cost and slows down the train, neither of which are good from a Patronage perspective.
From a purely Coverage point of view, of course, the people in these towns do need and deserve a lifeline of public transport. I am very much in favour of this, and will advocate for it - but it shouldn't be a train. Milk runs like that are the domain of the bus.
My ideal situation is essentially what happened when trains returned to Ararat in 2004. In the decade or so when no trains ran to Ararat, there were buses that served a lot of tiny towns between Ballarat and Ararat, like Trawalla and Buangor. When trains returned, it made no sense to stop the train at those places - so they kept the buses as well, and now there are a few trains per day and the gaps in the timetable are filled by buses. This adds up to a reasonable frequency for the bigger towns of Beaufort and Ararat, and doesn't leave the smaller ones stranded. Win-win.