Dumping Your Tree - A Radical Way To Correct Mistakes

Have you heard about the movement to abandon your family tree and start over again?  I first read about this trend last month as some genealogists decided to make this their New Year's resolution.
My first thought was, "Are you kidding me!?!  All these years of work abandoned to start over!  No, thanks."
The reasoning behind the idea is that many of us, when newbies, were happy clickers and not really evidence investigators.  By happy clickers I mean whenever I found info on someone's tree I would trust its accuracy and include it in my tree without ever analyzing it.  Thus, wrongly added info perpetuated errors as others copied it.
I know for fact this happened as recently I was investigating a collateral line for my Kinship Determination Project and uncovered an error that I've now found to have been copied by many others. Oops!  Although the error was innocent it really drives home the point that genealogists need to be careful and not rush.
I had found the name "Catherine" with a family in the 1860 census and assumed that was a child of the couple.  There were two genealogies written into book form from 1947 and 1959 but neither listed the child.  I figured they had just overlooked her as they had other children missing who had died in between census years that I had found via Find-A-Grave and Baptism records.
Little Miss Catharine grew up and I found her in the 1870 census not living with the family but attending a boarding school in the state the family had just relocated from.  That made sense to me, she was studying to become a teacher like her siblings.
In the 1880 census I found her married and living in the same town as "her parents."  In fact, Miss Catharine had married the widower of her sister who had died in 1879.
I found that couple in the 1900 census living in the same county and included in the household was who I thought was Catharine's father.  That was good enough for me.  Except, none of it was right!
Now that more records are available I found the marriage certificate for Miss Catharine and discovered her maiden name was not the family's surname.  So I looked for another marriage certificate, in several states, to see if Miss Catharine was also a widow and was using her married name and not her maiden name on the document.  Couldn't find one. The certificate did say it was her first marriage and the husband's second.  I had the husband's first marriage certificate so that confirmed his number of marriages.  I figured that recording it was her first marriage was an error but it was not.
The error was made in the 1860 census.  Upon closer examination I discovered that the enumerator had written Catharine but should have written Laura.  Catharine's year of birth is off by 5 years from the family's real child, Laura.  No, the names aren't close at all.  The mother's name was Catharine so I believe now that the record lists the mom twice and omits the daughter's name.  The Catharine in 1870 was a cousin of the family but not their child.  The cousin remained in the other state and married there. Have the marriage certificate to prove that and she is listed in the two genealogy books. The Catharine that was married to the widower was just another woman who happened to have the same name; she was not related in anyway to the original couple.  Now the 1900 census is very interesting in that the father of Laura is living with his ex-son-in-law and the new wife, Catharine, next door to his daughter.  The son-in-law was quite prosperous for the area, the couple had only 2 children and a servant living with them so they had plenty of room for the elderly man that his own children, with their large families living near by, did not.  So, I've corrected my error; I removed Catharine as their child. Interestingly, one of Catharine's children married into the kinship family so there is a connection, just not where I had it.
I do understand that as we improve our work that we will find errors, most likely many errors, that were made earlier in our career.  I'm still not sure that dumping your work is the solution.  I'm more apt to leave what I have and then go back and investigate closer line by line to make needed corrections.
As more direct evidence becomes available, past analysis may prove to be in error.  I'm okay with that!  I'd rather spend the time analyzing what I've already found then having to accumulate documents all over again. Now I'm working out a method to make sure I am able to go over my existing lines.  I wish I could color code or date stamp when I've touched a family so I know they've passed my review.  Since that doesn't exist, I've created an Excel document that has the family name, for example, Joseph Kos, a column for the date I began to check and a column for the date I've finished reviewing his line to where they connect to a living relative as I know that with my more recent family members, the information I've recorded is correct.  I've also created a spreadsheet called, Interesting Folks, and I'm listing the ancestors' name, year of birth and death, the fact that's interesting, and the area that's interesting.  For example, Joseph Kos' fact would be that he died young due to the Spanish Influenza epidemic.  The area would be medical.  This way I can quickly find some of the interesting family stories that get lost in the tree.  This method is basically creating an index to the tree and I just wish I had thought of it when I first began!  Happy Hunting!