How to get system back after grub fail

[J.V.]

I installed CentOS 6 with the default settings (sda1, sda2), read the 
grub manual and went to the root command prompt
and typed:

#grub-install /dev/sd1

/dev/sda1 was what was in the docs.

After that I reboot and get a black screen with a blinking white cursor.

I boot with the CentOS CD in rescue mode and get to the root prompt and 
type:

At the shell, I type:

#grub-install /dev/sda

I get a message:

/sbin/grub: Not found.
sh-3.2#

Why does this happen?

#which grub-install reveals:
/usr/sbin/grub-install

Where do I checkout the grub code to fix this incorrect path to 
/sbin/grub in /usrs/sbin/grub-install?

My last question is what do I do now specifically such that I can undo 
what grub-install did (very step by step specifically) so that my CentOS 
system will boot.

The disk in the drive was the only disk there when I installed CentOS, 
no other os was there.

Can anyone, anywhere tell me what exactly/specifically needs to be done 
to undo whatever it was that my first #grub-install /dev/sda1 messed up?

Is there a way to modify the grub code such that it never messes up a 
system to make it such that it does not boot?  like for example if 
something is wrong with a configuration or the command is not correct, 
it will not run and say "dave I can't do that" or something such that 
many hours are not sucked out of lives trying to get a system going again?

I am willing to go make grub more user friendly, apparently it needs 
some work if it will completely hose a customers system such that it 
will not boot.  Also I can edit the documentation with a list of things 
to not do , if indeed I did something wrong.

I just want my system to boot again and to use grub to boot off of one 
of three drives in the system.  Is this possible?

J.V.