instead, using just 1 transistor but 2 diodes (to simulate OR) would do the job?, this would save space on processors less silicon used?... but would the diode replacement introduce issues? (eg how many transistors could be saved on a CPU? but at what cost...)
Vs this method here..
(Attached), are diodes fast enough to replace the transistor though?
LOL - imagine a resistor between collector/emitter 
In general, it's too simplistic. Need transistor to pull output high, need transistor to pull output low, need other transistors to turn both off and let output go high impedance. And that's just 1 output pin.
You have not mentioned inputs, latching/clocking inputs in, etc.
Intel/AMD/TI/IBM, etc. use x-ray lithography to make really small transistors, squeezing millions & millions of them onto pinky-fingernail sized chips, and make tons of those chips at once on 8", 12", and maybe even larger silicon wafers.
The two circuits you have drawn are "non-restoring logic". This means the signal loses some of its gain and voltage at each stage. If a ripple counter were constructed from this or logic, it would fail. Each stage needs to have a gain of more than one. You stage has gain of less than one. Ten stages of a ripple counter would fall flat.
An inverter cannot be made from your stage. NAND or NOR are the two basic gates, because they can invert. OR gates lack the ability to invert.
Your logic works for one stage. Nice try.
AmbiLobe:
The two circuits you have drawn are "non-restoring logic". This means the signal loses some of its gain and voltage at each stage. If a ripple counter were constructed from this or logic, it would fail. Each stage needs to have a gain of more than one. You stage has gain of less than one. Ten stages of a ripple counter would fall flat.An inverter cannot be made from your stage. NAND or NOR are the two basic gates, because they can invert. OR gates lack the ability to invert.
Your logic works for one stage. Nice try.
One stage is all that's required for this simple circuit.
So which are you after:
"save space on processors less silicon used?"
or some other external circuit you want to implement?
In VLSI a diode may be bigger than a transistor... Don't assume these things scale down to the nano world!
What's your real question? What's the goal?
Regarding MarkT's point, I seem to recall that resistors are rarely used in IC design, instead some other semiconductor is used because a JFET connected Gate to Source as a constant current source is a -lot- smaller than the relatively large real estate taken up by a resistor of appreciable size.