I think what your telling me is that I must use transistors in that circuit, just to drive the LED's and carry that voltage over to the other side of the project, where it connects to 556 timer to produce a frequency output. There were no transistors mentioned in any of the project examples I found for what I'm doing. I can wire them in of course.
Actually, I'm not telling you anything other than the circuit is poorly designed with little or no consideration given toward the basic operating parameters of the IC employed. IF you wish to have a trouble free build, you must stay within the design parameters of the parts used in the design; that was the point I tried to make.
Voltage is NOT carried over anything. Simply put, voltage is dropped/reduced as a result of being applied across a load, resistance, an impendence, resulting in current flow. OR in other words, V = IR in terms of the voltage dropped across whatever load, simple or complex.
I see no other chips in the schematic you supplied so there is no way to evaluate your entire project at this time.
At 1ma, seems impossible. The circuit I have on my solderless breadboard powered by 5 volts. LED's and 1k resistors seems to work correctly, which would exceed the 1ma of current.
The 1ma typical @ Vdd=5V cited from the datasheet is the current the manufacturer guarantees as a source current for the outputs with criteria. It's not the ABSOLUTE limit but rather the SAFE and GUARANTEED current for every device produced WITHIN the boundaries described. You'll have to take that issue up with RCA who developed the CD 4000 series back in the late 1960's.
You say your components "seems to work correctly", but what have you DONE to evaluate their operational status? I have done a simulation of one output of your circuit with LT Spice, of which I'll attach the circuit as tested and modeled by the program using the spice models on hand along with a stack of three graphs showing the output current when the output goes high displaying the source current of that simple series circuit and the voltage at the two relevant points.
The current when the output is sourcing is ~2.21ma far in excess of the spec'ed 1ma @ 25deg C. Voh of Q0 is ~3.89V well below the spec'ed 4.6V for Vdd=5v at the spec'ed 1ma Ioh. The voltage drop across R6 is ~1.12V. See the TI 2004 datasheet page 3 for verification of that data. (http://www.ti.com/lit/ds/symlink/cd4017b.pdf)
I wasn't able to verify the part ID of the LED's on your schematic because the were far to small, but I figured from the 510 ohm limit R you were going for about 10ma for the LED's with the 5V Vdd. Because your LED's had some illumination with just over 2ma of drive is not at all surprising, but take one of those same LED's connect the anode to 5V in series with one of your 510 ohm R's to ground and compare the intensity of the LED with those driven by the 4017. You will see a big difference
Bottom line is you need to interface the counter with a transistor array, line driver or buffer that can handle the current. The cheap, easy solution is a 74HC244 octal buffer for your 8 outputs. It's about $0.50 a copy here in the States at either Mouser or Digi-Key by the unit. It is capable of driving a given line at +/- 35ma. If you go that route, be sure to deal with BOTH enables which are active LOW.
I hope this clears up some of your confusion regarding the importance of staying well within the manufacture's spec'ed criteria.