Lift Ships, a short story of possibilities: Copyright © bf, 2009-2014, All Rights Reserved

Take a train going down a valley and set up sensors in a circle. Each sensor station has two sensor sets, one to measure the speed of the light coming from headlight, the other to measure the speed of the train.
Every sensor station, except two, show that the speed of light is a constant no matter how fast the train is going.
The sensor looking at the caboose, sees the speed of the train, but can't see the headlight.
The sensor looking directly at the train sees the train and headlight and measures the speed of both, except at time 0/0 when the train hits the sensor and destroys it, resulting in no measurement.

Change the train to a bike with a headlight...The same measurements are taken and no reading is found when looking at the caboose end,  At time 0/0, the bike hits the sensor and tips over...

Silly, maybe, but real measurements are real measurements and should not be discarded just because they don't fit the preconceived notions ( the Galileo "defense").

The same analysis parameters can be used on modern scientific records, we discard the data that doesn't fit the design parameters. Everything is tweaked to fit standard curves, best straight lines or bell curves.

Computer recording systems may capture all of the raw data, but routinely discard discrepant data so that it all fits, before displaying it.

If you look at a set of raw data, before it is adjusted, you may find some data points that are completely out of line. These would typically be seen as spikes in the data.

The argument can be made that these are not real data points but errors caused by bad sampling, bumping a machine or gamma rays hitting a sensor or some other error factor. All of these arguments are valid.
If the test is run again, the discrepants may or may not show up a 2nd time, but after 5 or 10 iterations, it is quite likely that discrepants will re-appear, not necessarily in the same place, but as recordable information.
The question is are these are sampling errors or valid readings?  What are they saying to the researcher?

By accepting the discrepants and trying to fit them into a straight line or best fit, the lines become circular, spinning around the standard curve.

What is the mathematical formula to describe the new data? I have no idea (math tends to cause problems in my life).

The modified Einstein equation e=m(c does describe the output, but derivation of the formula is yet to be started.

Mass makes a difference:

Let's go back to the train sampling experiment:  Take measures 360' around the light source and 360' above and below the source..

Your assistant is handling the sensors in real time.

Oops..can't measure 180' below the light source for the train or those measurements become discrepant.

At 180' behind the train, no measurement.
At 180' in front of the train, time 0/0, no measurement: train+Assistant = squished Assistant
Likewise, at 180' behind the bike, no measurement, big bike butts are in the way of the light source.
At 180' in front of the bike, time 0/0, no measurement: Bike+Assistant=2 injuries

Now measure a Gnat, a full 360' around and above the insect is possible, if it stays still.

With a little training, in a couple of years, you can teach a Gnat to fly in a straight line down the valley. 

Attach a tiny helmet, with some cute, tiny ear flags and a LED light a few molecules thick.

Take readings above and around the light source; a full 360' 

No readings of the light are possible when looking at the Gnat's ass. 

As the Gnat approaches the assistant and at time 0.00001 / 0 the Gnat flies over the Assistant's head, because Gnats are not as stupid as trains, bicyclists or Assistants.

At time 0+0.1 The Assistant can now only measure the Gnat's ass....which, as we all know,  is where the derivation of the phrase "I don't give a Gnats ass" comes from.

So how does this relate to Einstein's seminal equation?

There are points in time and space that can't be measured leaving holes in the equations.

Using two overlapping "C's" spinning in opposite directions instead of C^2 indicates a dynamic measurement of the speed of light and how light can act as both a wave and a particle.