Thank you both for your interesting replies. I am particularly interested to learn from Prissi that there are provisions for a player rating: that seems as if it might be most worthwhile. Possible uses for player ratings include passengers using it as a weighting factor when deciding whether to use one or another transport company when both provide routes to their destination, or passengers using it as a weighting factor to decide whether or not to use private cars or not. It could also be a valuable tool in scenarios. One possible difficulty with going to far in loading times at crowded stations, however, is that, if it takes too long to load vehicles in a crowded station, there may come a point in which it is not possible to load passengers/goods onto the vehicles faster than they accumulate, leading to irreversible overcrowding. For this reason, any time penalty must be fairly modest, and cannot be sufficient in itself to be the only limiting factor on overcrowding.
The idea of trains taking longer to load at crowded stations is certainly a good one: there might well be some scope for having variable loading times for different sorts of vehicles in any event (another way, in effect, of distinguishing between long-distance and local vehicles, and an area in which vehicles can become more technically advanced with time), and crowding would be a useful factor to add to that.
As to capacity limitation: like other features of Simutrans, the purpose is to provide an abstracted representation of the aspects of reality that they are simulating and at the same time balance the gameplay. In real life, passengers will not wait weeks on end at a station so overcrowded that the first hundred trains that come in are too full to take any more passengers - people would simply leave and either find another mode of transport or abandon their journeys entirely. Perishable goods would perish if they had to wait too long to be transported, and would not sit around for weeks in the station and then pay differential rates depending on whether the journey took one or two hours once they were eventually loaded. So, whilst it should be quite possible for stations to become overcrowded, there should be some limit to how long that goods/passengers will wait and accumulate.
In respect of rates, it seems to me somewhat contradictory to have a speed bonus system in which passengers and cargo pay more or less depending on the speed of their journey, but then ignore the time spent waiting at an overcrowded station in the calculation of that amount. The idea of an overcrowding penalty for revenue would be to bring the performance of stops in line with the performance of vehicles in terms of revenue calculation, and also provide a strong financial disincentive to players to allow overcrowded stops: at present, there is no such disincentive.
Finally, to answer Combuijs' point, the solution to that in respect of public service stations would be to determine responsibility for overcrowding by checking the destinations of the passengers/cargo against the destinations served by lines serving the station: any company with lines serving that station with destinations including the destination of passengers whose numbers are so great as to cause overcrowding is partly responsible for that overcrowding.