An interesting Pakistani author (possibly of Indian origin from his writing) by the name of Abdul Jamil Khan has written a book called "Urdu-Hindi: an artificial divide" which has gained plenty of popularity amongst Indians.
As people know, anything about Pakistan that is dismissed as "Indian" or even linked as "Indian" gains popularity very fast in India or anything from a different South Asian country which is thrown under the "Indian" label causes joy amongst Indians, especially the expansionist ones.
I haven't read the book, save for a few pages from the first chapter, so I cannot comment on the entire book. However, everywhere I search this book on the internet, I read the same description of the points he makes in his book:
• Hindi evolved not from Aryan Sanskrit but from Pre-Aryan Dravidian and Austric-Munda rooted in Middle-East/Mesopotamia;
• Hindi’s script evolved from Aramaic system of writing similar to Greek;
• Urdu has not evolved as an offshoot of a prototype Indo-European language but it has its roots in Mesopotamian and Sumerian civilization;
• The ‘myth’ of Indo-European family of languages was created by Germans to satisfy their own theory of a superior German/Aryan race;
• Ancient Africa served as the melting pot of languages.
I'd like to give my personal views on each point of his hypothesis (however crazy it may sound) based on all the knowledge I have on the subject of linguistics which I'm well read on.
*On his first point regarding Hindi being of non-Indo-European origins, I cannot personally counter that claim since I am no qualified linguist, however based on the knowledge I have so far, I can see no evidence to this claim.
Linguists classify Hindi & Urdu as Indo-European languages because of their grammatical structure which counts more than anything else.
Both these languages display the use of gender like most Indo-European languages and are non-agglutinative, which means they do not represent actions or ideas in shortened suffixes (and prefixes in the case of certain agglutinative languages).
Mr Jamil Khan's hypothesis uses hardly any grammatical evidence, but instead simply what appears to be loanwords, borrowed consonants particularly from Dravidian languages.
He also mentions Hindi & Urdu being Subject-Object-Verb (SOV) languages, which is a trait common to most Indo-European languages.
This only goes against his claims of trying to link the two languages to Mesopotamian or Dravidian languages, which are definitely agglutinative, not SOV.
I agree with his 1.3 classification section regarding myths surrounding languages, however what he seems to be doing is no different, except by pushing in a different direction of his own.
Since he hardly uses any grammatical evidence, his theory has really has no basis.
*The claim that Hindi has it's script borrowed from Aramaic is possible according to many, since many Alphabets worldwide seem to be based on scripture used in the middle east, such as the Phoenician writing system.
However, Mr Jamil Khan is not the first to claim this.
*The author claims the Indo-European (IE) hypothesis was started of in Nazi Germany. This is an outright lie.
Hitler's belief of the Aryan race was promoted during the rise of the Nazi party in the early 19th century, whereas the IE family was proposed as early as the 18th century.
The first proposal for the existence of a common Indo-European was by Sir William Jones
Mr Jamil Khan then touches on Sir Willaim Jones's proposal, which contradicts his later statements of Nazi Germany having been the founder of Pan-Aryan nationalism.

*On ancient Africa being the 'melting pot' of languages, there's mistakes and truths on that. Firstly Africa is a diverse place when it comes to languages. Various, unrelated language families there exist as is for most continents of the world save for maybe Europe which is almost entirely Indo-European speaking.
'Melting pot' usually means cultures and races coming together and combining into new races and cultures. I do not know what the author means in this context. If he means many languages breeding into single languages, that is not always possible. Languages that are not genetically related unless they have corresponding structures which are entirely coincidental.
Outside of these points, I'd like to draw on is the other bias/inaccuracy/serious flaw I see in his material that I've read so far:
-The claim of Urdu & Hindi being the 'natural' common language of the subcontinent and Dividing Hindustani somehow divided the populations. Never in the recorded history of the subcontinent was there a single language people spoke. Even today India is so diverse linguistically (Pakistan too, but to a much lesser degree), so dividing a single language such as Hindustani into Urdu and Hindi can hardly have an impact on the majority of the people since neither Hindi nor Urdu was their first language.
The use of both languages as common forms of communication around Afghanistan, Pakistan, India, other parts of South Asia and parts of the Arabian/Persian Gulf region is a recent practice.
At no time before the Eighteenth century was Urdu used in the region of Pakistan or other parts of South Asia as a lingua Franca.
In short how can you use one or two commonly used languages to represent the people of an entire diverse region?
The Europeans previously used French to reach a common level of understanding one another a few centuries back and now they use English. Can either one or both of these languages be used to represent Europeans on linguistic, cultural or racial grounds as a whole?
As stated before, I read the first few pages of his book here .
A small note to add is that the overuse of Hindi in India and Urdu/Undri in Pakistan as a common language has caused rebellion in the past and still does. When I write overuse, I mean to say that these two languages are being enforced on those who do not speak it as a first language, at the expanse of their native languages.
So using these two or any two languages to represent a diverse region as the subcontinent cannot be taken seriously.
-More pro-Indian bias can be seen at his mention of the mythical "partition of India" already disproven here.
-The author mentions Aramaic to be the "mother" of Arabic, which no other linguist to my knowledge claims. According to linguistic knowledge, Aramaic is a cousin to Arabic, both being derived from a Proto-Semitic language. Refer to the chart below:

-His claim for Semitic & Persian influences on Urdu/Undri being 'ingredients' of the language further discredits his hypothesis on linguistic lines, since borrowed words in a language are not included in it's structural study in the process of it's classification.
More so, borrowed words are a frequent occurrence in almost every language known. Urdu or any other language can randomly borrow words from related or unrelated languages to enhance it's vocabulary; provided there is no grammar conflict with the language and the foreign/alien word it is borrowing.
That being said, all the Arabic words in Urdu are borrowed free of conflict with Urdu's Indo-European grammar and can easily be replaced with words from almost any other language family, again, provided there is no grammar conflict.
Examples are Turkish and Farsi borrowing from Arabic. Plenty of Arabic loanwords can be found in both these languages.
Yet they are considered unrelated because their grammatical structures differ. Farsi belongs to the Indo-European family (like Hindi & Urdu) as Arabic belongs to the Afro-Asiatic family, as Turkish belongs to the Altaic family.
And yet no one who screams of Urdu/Undri being a "mixed language" seems to take this into consideration.
By the logic of the author the modern English language has roots in Japan and he middle east because of loanwords like 'tycoon' from Japanese or 'bazaar' from Arabic.
-The claim that only about 7-10% of Urdu's vocabulary is derived from Sanskrit is irrelevant, since linguistics does not allow classification of languages based on their vocabulary, but rather the grammar they are laid out in.
Not only that, but such kind of classification as done by Jamil Khan, goes completely against the rules of the scientific study of languages themselves.
Languages with common sounding words are proposed relatives, but if you don't have corresponding grammar, the words are accepted as borrowed and/or coincidental. Languages with similar grammar, but lacking common root words and having distinct geography are regarded as having coincidental similarities; hence unrelated to each other.
Other observations that I find in his book which are interesting but hardly credible. In his first chapter where the author tries to 'prove' a Mesopotamian origin for Urdu & Hindi, he also tries to draw what seems to be genetic similarities between Indo-European, Afro-Asiatic, Altaic, Uralic & Dravidian.
Possible genetic similarities between these language families have been proposed by linguists before, which has led to the Nostratic theory .
If there is any credibility in this theory, it still won't prove Mr Khan's claims of Urdu & Hindi's "Dravidian, Mesopotamian roots", but rather a possible common origin between Dravidian & Indo-European languages alongside Semitic, which is a part of the Afro-Asiatic language family:

Below is a basic chart of the proposed Nostratic family:
