(Note: @pevaquark did some light editing aiming to make a few paragraphs to see where one idea stops and the next one starts)
In actuality there is not neccessarrily 50 mutations per generation, in fact its a yearly progression of depending on circukstances .1 to 1 bp per year so 25 year old having offspring would be between 2.5 and 25 bp mutations however only approximately 1 in 2 chance of those being passed at best however correction lowers that to 1 in 4 , remember chromosones are in sets of 2 and you only pass one so mutation on non paased sequence is meaningless so on average less than half of mutations are passed to offspring due to correction. Secondarily even the number of 50 mutations already takes in to account both parents and average age. So 50 per generation witha persistance rate of 50 percent into the following generation which again perpetuatis itself. Incomplete math is a pet peve of mine. 300000 generations and 50 per would yield only 15 million events in a lineage remember both parents are considered and the larger the population the less likely a full adoption of genomic change is to take place. And if we are talking distinct species a bottleneck event has to take place to seperate it. Now with the persistance generation to generation you can figure at most 7.5 of those 15 million to make it through the lineage if its a small population and vastly less if its a larger population simce more unaltered pairs exist to fix the error expecially with non advantageous pairings. This does not yield under any circumstance a 40 million base pair change over 7.5 million years. Alot of our genomic change expectations are based on currently observed mutation rates which are in a much more mutagenic environment with many more dietary variables and airborne toxins than would have been present at the supposid time of "evolution’ people who eat healthier less processed food and live in cleaner environments average closer to the .1 bp per year mark remember its a continuum not a constant. Also succesfull mutations on active sequences are almost always fatal and there are many points of transition between the species that would not have been servivable.
Also there is evedince lf more human like footprints pre dating some of th supposid common ancestors. Most of the radioactive dating methods have lots of evidence proving them unreliable. Radiometric dating for pompei as an example puts it at several million years old. And mount saint helens as well the valley created by it dates at over 2 million years. We have alot of chronology problems we need to fix obviously with those 2 there is an order and not only age issue since most tests show the more recent event to pre date the older one. This is a problem with many of the radiometric tests even tests of different parts of the same strata come up with different ages drastically different even we know they were formed together.
More of the differences are in active dna sequences and when you rate the probability of change in a certain dna sequence and work it out there is an even more definitive andswer as to simple random genetic mutation not being possible in the time frame. A speckfic utterance of the so called common ancestor equal distant from both prevailing species has not been found only proposed due to a mathmatical calculation based on the differences between the species. It for all the instances i have directly looked into has had many major errors in the calculation methodologies not sound genetically speaking. Propogation of genomic change throughout generations cannot be boiled down as simply as many mathmaticians have pre supposed. I like how einstein put it of all that is true of mathmatics nothing is true of reality.