1. You haven't said whether you are using the CMOS or bipolar version.
2. Your schematic shows 98 ohms for R1.
That means when your discharge pin 7 is "low", it must sink 50 milliamps through R1 plus some more from R2.
The IC should do that, but do not think pin 7 will go down to zero volts while sinking that much current.
So - if pin 7 is at a fraction of a volt instead of zero it'll slow down your discharge, which affects timing..
Take look at the spec sheets.
Look here for cmos version
http://www.ti.com/lit/ds/symlink/tlc555-q1.pdf
page 5 of 15 , specifications with 5 volt supply
parameter "Discharge-switch on-state voltage" about middle of page
it's allowed half a volt at just 10 milliamps.
Next, page 6 shows that with 15 volt supply it's allowed 1.7 volts at 100 milliamps
see page 8 fig 2, chart of "Discharge switch on state resistance" and observe it's probably around 10 ohms at 50 ma. That's a half volt.Look here for bipolar version
http://www.ti.com/lit/ds/symlink/ne555.pdf
page 5 0f 36, parameter "DISCH switch on state voltage" about middle of page
it's allowed
4/
10 volt at just 8 milliamps.
Now look at page 7 figure 1 , the curve of output volts versus current with 5V supply . Assuming the output and discharge transistors are similar, at 50 milliamps you might see more than a whole volt.
I think that's the source of your discrepancy.
Try making your discharge pin currents smaller by using larger R1 and R2.
Then the device will do a better job of discharging C1 so you should find yourself operating closer to the formulas.
