Accuracy of Regression Results?

#1
I'm new to, really, all statistical models and have done some fair research specifically on regression models. I have come across some questions that I seem to have a hard time finding a clear answer to. When doing a regression model (I'm using Excels built in data analysis tool) I am curious how accurate it may be compared to a regression from R-Statistics. If not much is different or should be the same, where do I move going forward with the data received from the regression report? i.e. What should be done with variables that have low p-values? Thanks.
 
#2
I haven't used the data analysis in excel, but you would look at residual plots to check for nonconstant variance, normally distributed error terms and other model adequacy checks.
 
#3
Excel regression is accurate ...it shows 5-6 digits which are correct.
Generally, there is no accuracy problem in regression's F statistic calculation, only matrix calculations and for p-value the F distribution in excel is accurate.