Let's assume we have a modulus in RSA with 2048 bits that was produced by multiplying two primes. What is the difference in digits in those two factors? For example, one prime is 1024 bits long and the second one has 1024 as well. Or one is 900 bits and the second one 1148 bits?
Are there any limits for primes that could be used in RSA or could it work with two primes where one prime consists of 1 bit and the other is a 2048-bit prime?
As far as the standard (PKCS#1) is concerned, the primes need not have the same length; the smallest should not be "too small" to avoid getting in range of the ECM factorization method, but there is still a lot of margin here.
However, some implementations can be more limited. For instance, Windows' implementation of RSA (in CryptoAPI) assumes that, for a 2048-bit modulus, both primes have length exactly 1024 bits, no more. It is also commonplace in hardware-assisted implementations (think smart card) to enforce such a constraint, because they do computations modulo p and modulo q (with the Chinese Remainder Theorem) and thus need circuitry able to process numbers up to the size of the larger of the two primes -- they thus have a vested interest in making the larger prime as small as possible, leading to both primes having the same length.
RSA supports moduli that are a product of more than two primes, but, there again, it must not be overdone, and support for the encoding of private keys with more than two primes is not widespread. This does not matter when private keys are not exported. When you use the public key, you don't know, and do not have to know, the number and sizes of the prime factors.
When you multiply an $n$-bit integer by an $m$-bit integer, the product is an $(m+n-\epsilon)$-bit integer, with $\epsilon \leq 1$.
For RSA the security recommendations specify that secret factors of the public modulus must be of approximately the same size. If the difference in the number of bits is modest (less than a machine granularity) it could be approved.
There is also another suggestion from Shamir about generation of RSA keys for paranoid, which recommends for example a prime $p$ of say $1000$ bits and the other prime $q$ of $5000$ bits.
To get a sense for common prime lengths in RSA, you can run
openssl genpkey -algorithm RSA -out private.pem
to generate a private RSA key. Then, have a look at the primes with
openssl rsa -in private.pem -text -noout
I'm using OpenSSL 1.1.1m and both primes are 2048-bit. The public modulus has 4096 bits.