The problem with RF transmitter modules is that they are ultimately designed to be cheap and cheerful, at the basic level just an oscillator which can be turned on and off which generates a radio signal or not. The price one pays for low cost is put on the end user who has to do everything to control the module and work round its limitations.
As buffering, hand shaking and whatevery else gets added to make the end-user's life easier the cost of the modules rise because of component and R&D costs.
With commercial customers looking primarily at component costs ( their R&D costs being minimal compared to costs saved on large manufacturing runs ), it's hard to create a compelling product which gets them to pay the money and move in-house development costs over to compnent costs. More advanced, more easily usable modules, tend to therefore end up as more niche markets and don't have the economics of scale which brings their price down.
Move enough into a module and the balance can be shifted. The price of a complete module and what it does outweighs the in-house R&D.
What's really being asked for is XBee for 433MHz rather than for ZigBee, and at low cost.
Whether that will ever happen really depends on whether a manufacturer sees there's a large enough market to justify it. For those who want that, it can all be concocted using cheap microcontrollers, so manufacturers may see no advantage or market for doing it themselves.
Local CTS/RTS handshaking could probably be achieved by putting a PICAXE between the master and transmitter and it would increase the buffer size as well. It may well be possible to use cheaper, dumber TX modules as functionality moves into the intermediate PICAXE.
Whether people are prepared to do that, or pay the cost for it, is a different matter. Until a manufacturer does add it themselves the options seems to be not have it, pay the cost of working around what you have to work with, or move to a different technology which does provide what's wanted.